{ "run": "2026-10-08", "fixture": "ai-answer-check-v1", "calls_published": 70, "calls_dropped_exclusion": 0, "summary": { "aimode": { "calls": 10, "with_sources": 9, "fresh_correct": 3, "fresh_graded": 4, "web_search_reported_true": null, "returns_model": false, "returns_timestamp": false, "returns_search_queries": 0, "median_latency_ms": 10359, "cost_usd_total": null }, "chatgpt-api-nosearch": { "calls": 10, "with_sources": 0, "fresh_correct": 0, "fresh_graded": 4, "web_search_reported_true": 0, "returns_model": true, "returns_timestamp": true, "returns_search_queries": 0, "median_latency_ms": 6497, "cost_usd_total": 0.013387 }, "chatgpt-api-search": { "calls": 10, "with_sources": 6, "fresh_correct": 3, "fresh_graded": 4, "web_search_reported_true": 6, "returns_model": true, "returns_timestamp": true, "returns_search_queries": 6, "median_latency_ms": 5465, "cost_usd_total": 0.182684 }, "chatgpt-web": { "calls": 10, "with_sources": 10, "fresh_correct": 4, "fresh_graded": 4, "web_search_reported_true": null, "returns_model": false, "returns_timestamp": false, "returns_search_queries": 0, "median_latency_ms": 28893, "cost_usd_total": null }, "chatgpt-web-nosearch": { "calls": 10, "with_sources": 10, "fresh_correct": 4, "fresh_graded": 4, "web_search_reported_true": null, "returns_model": false, "returns_timestamp": false, "returns_search_queries": 0, "median_latency_ms": 29366, "cost_usd_total": null }, "gemini": { "calls": 10, "with_sources": 10, "fresh_correct": 3, "fresh_graded": 4, "web_search_reported_true": null, "returns_model": false, "returns_timestamp": false, "returns_search_queries": 0, "median_latency_ms": 42302, "cost_usd_total": null }, "perplexity-routed": { "calls": 10, "with_sources": 10, "fresh_correct": 4, "fresh_graded": 4, "web_search_reported_true": 10, "returns_model": true, "returns_timestamp": true, "returns_search_queries": 0, "median_latency_ms": 5334, "cost_usd_total": 0.060269 } }, "panel_brand_overlap": { "P21": { "chatgpt_web": [ "Ahrefs", "DataForSEO", "Search Console", "Semrush", "SerpApi" ], "chatgpt_api_nosearch": [ "Ahrefs", "Majestic", "Moz", "Search Console", "Semrush" ], "shared": [ "Ahrefs", "Search Console", "Semrush" ] }, "P22": { "chatgpt_web": [ "Ahrefs", "Moz", "Search Console", "Semrush" ], "chatgpt_api_nosearch": [ "Ahrefs", "Moz", "Search Console", "Semrush", "Ubersuggest" ], "shared": [ "Ahrefs", "Moz", "Search Console", "Semrush" ] }, "P23": { "chatgpt_web": [ "DataForSEO" ], "chatgpt_api_nosearch": [ "Bright Data", "DataForSEO", "SerpApi", "Zenserp" ], "shared": [ "DataForSEO" ] }, "P25": { "chatgpt_web": [ "Ahrefs", "DataForSEO", "Keywords Everywhere", "Search Console", "Semrush" ], "chatgpt_api_nosearch": [ "Ahrefs", "Moz", "Search Console", "Semrush" ], "shared": [ "Ahrefs", "Search Console", "Semrush" ] }, "P02": { "chatgpt_web": [ "Apollo", "Crustdata", "Hunter", "People Data Labs", "ZoomInfo" ], "chatgpt_api_nosearch": [ "Apify", "Clearbit", "Crunchbase", "Hunter", "ZoomInfo" ], "shared": [ "Hunter", "ZoomInfo" ] }, "P26": { "chatgpt_web": [ "Ads Transparency", "Apify", "Meta Ad Library", "ScrapeCreators", "SerpApi" ], "chatgpt_api_nosearch": [ "Ads Transparency", "BigSpy", "Semrush" ], "shared": [ "Ads Transparency" ] } }, "calls": [ { "id": "F1", "route": "aimode", "latency_ms": 9845, "fetched_at": "2026-10-07T22:34:15Z", "answer": "The multimodal filter in Google Search Console is a performance report filter designed to isolate and track how your website's content appears in search results when users initiate a search using an image rather than text.\n\nAnnounced by Google in September 2026, this feature helps site owners measure visual discovery and understand how media-rich pages perform across modern Google search experiences.\n\nWhat Traffic Does It Include?\n\nWhen you apply the multimodal filter under the Web search type, it aggregates data from visual and mixed-input interactions, including:\n\nGoogle Lens\n\nCircle to Search on Android devices\n\nImage uploads directly into Google Search\n\nChrome's right-click \"Search this image\" feature\n\nWhere Can You Find It?\n\nAvailable inside the standard Performance on Search results report (under the Search type filter).\n\nIntegrated into the Generative AI features report.\n\nFully exportable if you want to analyze visual search performance metrics alongside your other data tools.\n\nKey Limitations to Keep in Mind\n\nNo text query data: Because these searches start with a visual asset rather than a typed keyword string, the Queries tab is disabled or unavailable when the multimodal filter is active.\n\nData availability: Metrics will only display for properties that actively receive impressions or clicks originating from these image-based search features.\n\nYou can read the official rollout announcement on the Google Search Central Blog or learn more about the breakdown via Search Engine Journal's coverage.", "sources": [ "https://developers.google.com/search/blog/2026/09/web-multimodal-in-sc", "https://www.searchenginejournal.com/google-search-console-multimodal-filter/590781/", "https://scalevise.com/resources/google-search-console-multimodal-filter-2/" ], "model": null, "provider_timestamp": null, "web_search": null, "search_queries": null, "served_by": null, "n_sources": 3, "grade": "correct", "brands": [ "Search Console" ] }, { "id": "F2", "route": "aimode", "latency_ms": 13687, "fetched_at": "2026-10-07T22:34:29Z", "answer": "No, Ahrefs did not make a general change to its core API rate limits or unit allocations in September 2026.\n\nAPI and Feature Changes around this Period\n\nWhile general API limits did not change in September 2026, Ahrefs introduced a related Brand Radar API timeline adjustment and had prior updates earlier in the year:\n\nBrand Radar Volume Update: The search_volume_type=keyword_volume option on Brand Radar API endpoints remained available through September 30, 2026, shifting all requests to AI-adjusted volume starting October 1, 2026.\n\nPrior API Adjustments: Higher overall API limits were rolled out earlier in the year as part of broader platform updates in April 2026, and free Domain Rating endpoint requirements were adjusted in August 2026.\n\nIf you are looking for specific rate limits or credit quotas for a particular Ahrefs API plan (e.g., V2 or Brand Radar), let me know and I can detail those exact thresholds for you.", "sources": [ "https://docs.ahrefs.com/en/api/docs/changelog", "https://ahrefs.com/blog/new-features-apr-2026/" ], "model": null, "provider_timestamp": null, "web_search": null, "search_queries": null, "served_by": null, "n_sources": 2, "grade": "wrong", "brands": [ "Ahrefs" ] }, { "id": "F3", "route": "aimode", "latency_ms": 9361, "fetched_at": "2026-10-07T22:34:38Z", "answer": "OpenSEO is a 100% open-source, self-hosted SEO platform designed to provide transparent, flexible, and cost-effective search engine optimization workflows without vendor lock-in. It is built with an AI-native approach\u2014featuring Model Context Protocol (MCP) support\u2014allowing you to connect it with AI agents (like Claude) for tasks like technical audits, keyword research, competitor analysis, and rank tracking.\n\nKey Features of OpenSEO\n\nSelf-Hosted & Open Source: Gives you full privacy, data control, and customizability.\n\nAI Agent Integration: Utilizes MCP commands (such as /seo-coach, /seo-audit, and /keyword-research) to let AI assistants research and run workflows directly.\n\nGoogle Search Console Integration: Connects with GSC to analyze your site's actual performance and queries.\n\nComprehensive Toolkit: Handles keyword research, competitor analysis, backlink tracking, and technical audits.\n\nData Source\n\nDataForSEO: OpenSEO relies on DataForSEO as its primary pay-by-usage data provider. This allows users to access reliable, large-scale historical SEO and SERP metrics while avoiding expensive subscription models or \"black box\" pricing.\n\nYou can explore the project directly on the OpenSEO Official Website.\n\nWould you like to know more about how to self-host OpenSEO or how its AI agent and MCP integration works?", "sources": [ "https://openseo.so/", "https://openseo.so/open-source-seo", "https://openseo.so/features/keyword-research" ], "model": null, "provider_timestamp": null, "web_search": null, "search_queries": null, "served_by": null, "n_sources": 3, "grade": "correct", "brands": [ "DataForSEO", "Search Console" ] }, { "id": "F4", "route": "aimode", "latency_ms": 8864, "fetched_at": "2026-10-07T22:34:47Z", "answer": "Jev is a frontier AI model developed by TypeSafe AI that is purpose-built to make fast, structured decisions instead of generating text.\n\nLaunched in September 2026, it was created by Diogo Almeida, a co-creator of ChatGPT and RLHF who previously worked at OpenAI. Rather than acting as a conversational chatbot, Jev operates as a \"smart if statement\" meant to be integrated directly inside software pipelines to process massive datasets, automate workflows, or serve as guardrails for other AI systems.\n\nKey Features and Architecture\n\nThe \"System One\" Approach: Named after Daniel Kahneman\u2019s concept of fast, intuitive human thinking, Jev acts on split-second judgment calls. It focuses entirely on immediate classification and evaluation, bypassing the heavy multi-step token reasoning of traditional large language models (LLMs).\n\nNo Text Generation (Zero Hallucinations): Jev cannot write prose, chat, or generate sentences. Because it is a discriminative model rather than a generative one, it cannot \"hallucinate\" an invalid answer. It only returns type-safe structured data (e.g., JSON).\n\nCalibrated Probabilities: Jev uses a specialized training method called Reinforcement Learning for Calibrated Decisions (RLCD). When answering a question, it outputs specific confidence metrics and real-world probability estimates alongside its choices, allowing developers to set exact mathematical thresholds for automation.\n\nExtreme Speed and Low Cost: Because it processes all target choices simultaneously without streaming text token-by-token, it runs in 70 to 500 milliseconds. It is highly cost-effective, priced at roughly $0.04 per million input tokens with zero output charges.\n\nHow Jev Evaluates Data\n\nInstead of standard conversational prompts, developers hand Jev a piece of \"state\" (raw text or JSON data) and evaluate it against three question types:\n\nChoice: Choosing the best-fitting option out of a fixed, pre-defined array (e.g., sorting an inbound support email into Refund, Technical Bug, or General Inquiry).\nScore: Evaluating the input state against a rigid rubric to assign a graded numeric value and probability.\nNoul: Returning a definitive binary response (Yes/No) paired with a 0 to 1 confidence scale.\n\nPrimary Use Cases\n\nEmail and Ticket Triage: Routing millions of incoming user requests to the correct software pipelines instantly.\n\nAgentic Guardrails: Inspecting tool calls made by other generative AI agents (like Claude or GPT) to block risky or unauthorized actions before they execute.\n\nHigh-Volume Content Moderation: Flagging, tagging, or categorizing massive streams of user content (like video comments or forum posts) in real time.\n\nWould you like to see a code example of how a developer implements a Choice or Noul query using Jev, or look into its integrations with frameworks like LangChain?", "sources": [ "https://typesafe.ai/blog/introducing-system-one-models-and-jev", "https://www.youtube.com/watch?v=YGgNBcIgI4s", "https://www.lennysnewsletter.com/p/jev-for-beginners-how-to-use-it-and", "https://www.langchain.com/blog/building-a-harness-with-jev", "https://www.youtube.com/watch?v=77owx_9zLvY", "https://www.youtube.com/watch?v=0zFfcEr1e9U", "https://www.mindstudio.ai/blog/jev-system-one-model-launch", "https://www.sanity.io/glossary/jev-typesafe-ai-model", "https://www.youtube.com/watch?v=ZgXej_9isxY", "https://zapier.com/blog/jev/", "https://en.wikipedia.org/wiki/Jev_(AI_model)", "https://beam.ai/agentic-insights/jev-typesafe-ai-agents", "https://greennode.ai/blog/what-is-jev" ], "model": null, "provider_timestamp": null, "web_search": null, "search_queries": null, "served_by": null, "n_sources": 13, "grade": "correct", "brands": [] }, { "id": "P02", "route": "aimode", "latency_ms": 11267, "fetched_at": "2026-10-07T22:35:36Z", "answer": "The cheapest way to enrich company domains without a monthly subscription is using pay-as-you-go credit APIs or leveraging generous free-tier allowances from B2B platforms.\n\nInstead of locking into rigid monthly SaaS fees, non-subscription workflows rely on purchasing non-expiring or pre-paid credit bundles, or utilizing free monthly allotments for low-volume needs.\n\nTop Non-Subscription and Low-Cost Enrichment Options\n\nProvider / Method\nPricing Model\nBest Use Case\nPeople Data Labs (via on-demand brokers like Notis.ai)\nPay-per-call (~$0.12/call) with no account or monthly vendor bill\nAd-hoc domain lookups without committing to a contract.\nApollo.io\nFree Tier offering robust monthly search and enrichment limits\nZero-budget workflows requiring steady, moderate-volume checks.\nOrbitly\nPay-as-you-go credit packs with zero monthly commitment\nAppending firmographic attributes intermittently.\nEnrich Layer\nPre-paid credit bundles that do not expire on a monthly drip-feed\nBursty, project-based data enrichment.\n\nAlternative DIY Scripting Method\n\nIf you have basic technical capabilities, you can build a near-free enrichment pipeline using custom web scraping:\n\nWrite a lightweight Python script using libraries like BeautifulSoup or requests to target the target company's domain home page and /about or /contact pages.\nExtract meta tags, copyright text, and social links directly from the raw HTML.\nThis reduces your direct cost to near-zero (only costing standard proxy or server compute fees if done at scale).\n\nIf you can share your estimated monthly volume of domains and what specific data points you need (e.g., employee count, tech stack, industry, revenue), I can recommend the exact provider tier that fits your budget.", "sources": [ "https://syncgtm.com/blog/best-free-enrichment-tools-2026", "https://www.notis.ai/blog/how-to-enrich-a-company-from-just-its-website/", "https://www.outflo.io/sales-tools/enrichlayer", "https://www.vanderbuild.co/sales-tools-database/tool/orbitly" ], "model": null, "provider_timestamp": null, "web_search": null, "search_queries": null, "served_by": null, "n_sources": 4, "grade": null, "brands": [ "Apollo", "People Data Labs" ] }, { "id": "P21", "route": "aimode", "latency_ms": 8361, "fetched_at": "2026-10-07T22:34:56Z", "answer": "For an in-house growth agent\u2014especially one building automated loops, internal growth models, or autonomous AI workflows\u2014choosing the right SEO API depends on whether you need raw, pay-as-you-go infrastructure to build custom logic or ready-made enterprise intelligence with massive proprietary indexes.\n\nThe top SEO data APIs are organized below by their ideal growth use case:\n\nDataForSEO \u2014 Best for pure developers & custom agentic workflows. Unlike traditional SaaS tools, it has no front-end dashboard\u2014it is 100% API-first. It provides modular, pay-as-you-go endpoints for SERP, keyword data, backlinks, and on-page metrics across multiple engines (Google, Bing, Amazon, YouTube). It also features native integrations for automation stacks like n8n and DataForSEO MCP setups.\n\nSE Ranking API \u2014 Best overall for hybrid project automation & AI visibility tracking. It offers a robust dual-architecture (Data API and Project API) that lets you harvest bulk metrics while automating rank tracking and site audits. SE Ranking stands out for modern growth stacks through native support for SE Ranking Model Context Protocol (MCP), allowing agents to plug directly into LLMs (like Claude or ChatGPT) without custom middleware.\n\nAhrefs API (v3) \u2014 Best for backlink intelligence & competitive link growth. Ahrefs operates one of the most active web crawlers and largest live backlink indexes on the internet. If your growth agent relies on link-building automation, authority scoring, or deep competitor backlink discovery, its programmatic endpoints in Site Explorer and Keywords Explorer are the industry gold standard.\n\nSerpApi \u2014 Best for real-time, multi-engine SERP scraping. If your growth agent needs literal, exact-match SERP HTML or structured JSON across Google, Bing, DuckDuckGo, or YouTube without managing proxies or captcha blocks, SerpApi handles the heavy lifting reliably on a flexible credit tier.\n\nGoogle Search Console (GSC) API \u2014 The mandatory baseline (Free). No in-house growth agent is complete without pulling first-party data directly from the Google Search Console API. It remains the absolute source of truth for your own property\u2019s actual impressions, clicks, average positions, and query distribution.\n\nTo help narrow down the best fit, tell me:\n\nIs your growth agent building a custom internal tool from scratch or integrating with existing dashboards?\nWhat specific data are you prioritizing (e.g., rank tracking, backlinks, AI-search visibility/citations, or SERP scraping)?", "sources": [ "https://www.nimbleway.com/blog/top-seo-analysis-apis", "https://www.position.digital/blog/best-seo-geo-api/" ], "model": null, "provider_timestamp": null, "web_search": null, "search_queries": null, "served_by": null, "n_sources": 2, "grade": null, "brands": [ "DataForSEO", "Ahrefs", "SerpApi", "SE Ranking", "Search Console" ] }, { "id": "P22", "route": "aimode", "latency_ms": 7281, "fetched_at": "2026-10-07T22:35:03Z", "answer": "Getting keyword volumes and backlink data without paying for multiple heavy monthly subscriptions (like Ahrefs, Semrush, or Moz) is totally doable. You just have to mix and match free tiers, alternative toolsets, and raw data extraction tricks.\n\nHere is how you can build a low-cost or free stack to get what you need:\n\n1. Finding Keyword Search Volumes (Without a Big SEO Sub)\n\nGoogle Keyword Planner: Free inside a Google Ads account (you don't even need to run a paid campaign to see broad search volume ranges and forecasts, though spending a tiny amount or having an active setup sometimes unlocks exact numbers).\n\nBrowser Extensions (Keywords Everywhere / WMS Everywhere): While Keywords Everywhere has a paid model now, it operates on a very cheap, pay-as-you-go credit system rather than a forced monthly subscription. Other free extensions like SEOquake or Detailed SEO pull some data on the fly.\n\nAnswerThePublic / AlsoAsked: Great for finding question-based keywords and content ideas for free (limited daily searches), helping you build out topic clusters without digging into massive keyword databases.\n\nGoogle Search Console (GSC): If you already have a site up, GSC gives you the exact impressions, clicks, and average positions for keywords you already rank for. It's real-world volume data you don't have to guess at.\n\n2. Checking Backlinks (Without Ahrefs/Semrush)\n\nGoogle Search Console: Under the \"Links\" report, you can see your top-linked pages, top linking sites, and anchor texts for free. It is the most accurate view of how Google sees your backlink profile.\n\nBing Webmaster Tools: Similar to GSC, Bing offers a solid, free backlink explorer tool that often catches links Google misses or indexes differently.\n\nHoth Backlink Checker / LinkMiner (Free Tiers): Tools like The HOTH Free Backlink Checker or Mangools' free tier let you run quick spot-checks on competitor domains to see a handful of their top linking root domains and authority scores without paying a cent.\n\nAhrefs Webmaster Tools (AWT): If you verify ownership of your own site via Ahrefs Webmaster Tools, Ahrefs lets you see all the organic keywords your own site ranks for and all the backlinks pointing to your site for free.\n\n3. Alternative & Hybrid Approaches\n\nGroup Buys / Shared Tools: Some independent marketers use SEO tool group-buy services (use with caution regarding data privacy and terms of service) to share access to high-end tools for a fraction of the cost.\n\nAPIs & Pay-As-You-Go: If you only need data occasionally, look at data providers like DataForSEO or a credits-based API where you only pay for the exact rows of data you pull rather than a $99/month blanket fee.\n\nIf you want to narrow this down, let me know:\n\nAre you checking your own website or analyzing competitors?\nWhat is your budget (fully free vs. occasional pay-as-you-go)?\n\nI can recommend the exact combination of tools for your workflow.", "sources": [], "model": null, "provider_timestamp": null, "web_search": null, "search_queries": null, "served_by": null, "n_sources": 0, "grade": null, "brands": [ "DataForSEO", "Ahrefs", "Semrush", "Moz", "Search Console", "Keywords Everywhere", "Mangools" ] }, { "id": "P23", "route": "aimode", "latency_ms": 10581, "fetched_at": "2026-10-07T22:35:14Z", "answer": "Running weekly keyword and SERP checks via an API as cheaply as possible comes down to avoiding high monthly subscription minimums and utilizing asynchronous batch processing.\n\nThe absolute cheapest and most scalable method depends on whether you are tracking existing impressions or pulling fresh SERP layouts.\n\nThe 3 Cheapest Approaches Ranked\n\nThe Free Route: Google Search Console (GSC) API\n\nCost: $0 (100% Free)\nBest For: Tracking keywords your domain already ranks for.\nHow it works: You can programmatically query the GSC API weekly to pull average positions, impressions, and clicks for your tracked queries.\nCatch: It only tracks data for keywords you already get impressions/clicks on. It won\u2019t show you competitor rankings or brand-new target keywords where you have zero visibility yet.\n\nThe King of Raw Scale: DataForSEO SERP API (Standard Queue)\n\nCost: $0.0006 per SERP ($0.60 per 1,000 requests) with a $50 minimum pay-as-you-go deposit.\nBest For: High-volume, programmatic rank tracking of any keyword or competitor.\nHow it works: Use their Standard Mode (asynchronous batch POST requests) rather than the synchronous Live mode. You drop up to 100 tasks in a single POST call, they process it in the background queue, and you pull the results.\nCatch: Standard queue isn\u2019t instantaneous (results take a little time to populate in the background queue), but it yields identical data to live/priority modes at a fraction of the cost.\n\nThe Zero-Commitment Alternative: Serpent API\n\nCost: Pay-as-you-go starting as low as $0.60 per 1,000 searches on default tiers, scaling down with higher usage, with no monthly subscriptions or minimum deposits.\nBest For: Low-to-medium volume checks where you don't want to tie up cash in a pre-funded wallet.\nHow it works: Simple pay-per-call endpoints for Google, Bing, and other search engines.\n\nCost Comparison Example (1,000 Keywords Checked Weekly = 4,000 Checks/Month)\n\nGoogle Search Console API: $0\n\nDataForSEO (Standard): 4,000 \u00d7 $0.0006 = $2.40 per month (plus the initial $50 account funding)\n\nSerpApi (Starter Tier): $25.00 per month (includes 1,000 searches; you\u2019d need a higher tier for 4k searches, making it significantly more expensive)\n\nArchitectural Tips to Keep Costs Rock-Bottom\n\nBatch Your Requests: If using DataForSEO, always bundle tasks into multi-item POST arrays instead of single ping-per-keyword live calls. Live calls cost more than double ( $0.002 vs $0.0006 per request).\n\nCache Results Locally: Since you are only checking weekly, store your JSON results in a lightweight database (like PostgreSQL or SQLite). Never re-query an API to display historical rankings on your dashboard; only pull fresh data on your designated cron-job day.\n\nFilter Geographies & Devices: Make sure your API payload explicitly defines desktop/mobile and country codes strictly as needed. Passing unnecessary or broad localization parameters can sometimes default to more expensive data centers or multi-region proxy fees depending on the provider.\n\nIf you can share:\n\nHow many total keywords you plan to track weekly\nWhether you need to track competitor domains or just your own\n\nI can calculate an exact monthly cost breakdown and help you structure a sample cron/batch script outline.", "sources": [ "https://dataforseo.com/pricing", "https://dataforseo.com/help-center/best-web-search-api-for-developers-what-to-look-for-and-why-dataforseo-serp-api-fits", "https://apiserpent.com/blog/dataforseo-pricing-explained", "https://answersocrates.com/blog/best-keyword-research-apis/", "https://www.rankability.com/blog/best-rank-tracking-apis/", "https://apiserpent.com/blog/serp-api-pricing-comparison", "https://serpapi.com/pricing" ], "model": null, "provider_timestamp": null, "web_search": null, "search_queries": null, "served_by": null, "n_sources": 7, "grade": null, "brands": [ "DataForSEO", "SerpApi", "Search Console" ] }, { "id": "P25", "route": "aimode", "latency_ms": 11098, "fetched_at": "2026-10-07T22:35:25Z", "answer": "Connecting Google Search Console (GSC) and keyword research data to Claude (including Claude Code and Claude Desktop) relies heavily on the Model Context Protocol (MCP) or structured data connectors. MCP allows Claude to natively query external APIs, execute tools, and pull live metrics directly into your workflow.\n\nThe best tools, connectors, and MCP servers available for bridging GSC and keyword research data with Claude include:\n\nmcp-gsc (Open-Source GSC MCP Server): An open-source, highly-rated MCP server specifically built for Google Search Console integrations. It supports over 20 tools covering search analytics, deep URL inspection, sitemap management, and property handling. It connects smoothly via local CLI environments or Docker, making it ideal for developer-heavy workflows like Claude Code.\n\nPorter Metrics MCP: Offers a streamlined custom connector URL (https://portermetrics.com) that you can plug directly into Claude\u2019s connector settings. It authenticates securely via your Google account, handles multi-property data blending automatically, and optionally hooks into BigQuery for massive data scales.\n\nCoupler.io (No-Code Data Pipeline): A robust no-code connector that processes GSC and Google Analytics 4 (GA4) data on its own servers and structures it as SQL/clean impression-weighted numbers before sending it to Claude. It is especially useful if you want to bypass Google's 16-month data window by auto-archiving historical metrics into a data warehouse like BigQuery.\n\nInsightful Pipe MCP: Tailored for generating executive-ready SEO audit reports, this integration aggregates 30-day GSC trends (clicks, impressions, CTR, average positions, mobile vs. desktop breakdown, and striking-distance keyword analysis) directly into structured Claude documents.\n\nGoogle Search Console MCP (via MCP Market): A specialized monitoring and analytics server (byahonn) designed to let AI clients request custom date ranges, query dimensions, and precise performance metrics on demand.\n\nIf you want to proceed, let me know:\n\nAre you looking for a local open-source CLI setup (like mcp-gsc) or a hosted connector (like Porter or Coupler)?\nDo you also need to blend Google Analytics 4 or keyword volume data (from Ahrefs/SEMrush APIs) into Claude?\n\nI can provide the exact configuration JSON or setup steps for your preferred path.", "sources": [ "https://modelcontextprotocol.io/docs/2026-07-28/getting-started/intro", "https://github.com/AminForou/mcp-gsc", "https://portermetrics.com/en/tutorial/claude/chat-google-search-console/", "https://www.youtube.com/watch?v=36J785eT1JM", "https://mcpmarket.com/server/google-search-console", "https://www.youtube.com/watch?v=Ff0qDyZJVMQ" ], "model": null, "provider_timestamp": null, "web_search": null, "search_queries": null, "served_by": null, "n_sources": 6, "grade": null, "brands": [ "Ahrefs", "Semrush", "Search Console" ] }, { "id": "P26", "route": "aimode", "latency_ms": 10359, "fetched_at": "2026-10-07T22:35:46Z", "answer": "When researching competitors across both Meta and Google ad libraries programmatically, you have to navigate official APIs (which come with high verification hurdles) versus third-party aggregators and scrapers (which offer unified, multi-platform endpoints).\n\nThe top options break down by official access, multi-channel scrapers, and developer-friendly alternatives:\n\nOfficial Meta Ad Library API \u2014 The sanctioned, free-to-use option directly from Meta. It requires business verification and a formal app review. While robust for Meta-specific data (Facebook, Instagram), it has steep approval friction, restrictive rate limits, and a setup time that can take weeks.\n\nAdLibrary.com API \u2014 A top multi-channel choice if you want to bypass managing separate API connections. It provides a single REST endpoint covering Meta, Google, TikTok, LinkedIn, and more using one API key, drastically cutting down time-to-first-query compared to juggling separate platform approvals.\n\nApify (Meta/Google Scraper Actors) \u2014 Excellent for data engineering pipelines. Apify hosts pre-built scrapers (\"actors\") that structure raw data from the ad libraries into clean JSON, CSV, or database sinks that you can trigger via their API without dealing with official developer app reviews.\n\nSearchAPI.io \u2014 Offers dedicated programmatic search endpoints for specific ad repositories (like Meta) on a pay-as-you-go, per-request pricing model, making it ideal for lightweight tracking without heavy monthly subscriptions.\n\nForeplay \u2014 Highly rated for creative intelligence and visual swipe files. While its API is more limited, it is an industry favorite for teams wanting structured competitor ad creative analysis rather than just raw JSON payloads.\n\nTo help narrow down the best fit, let me know:\n\nAre you building an automated analytics dashboard / internal pipeline or just looking for a frontend research tool?\nDo you also need coverage for TikTok, LinkedIn, or YouTube, or strictly Meta and Google?", "sources": [ "https://adlibrary.com/alternatives/ad-library-alternative-2026", "https://adrio.ai/best/meta-ad-library-api-alternatives", "https://adlibrary.com/guides/best-ad-spy-api" ], "model": null, "provider_timestamp": null, "web_search": null, "search_queries": null, "served_by": null, "n_sources": 3, "grade": null, "brands": [ "Apify", "SearchAPI", "Meta Ad Library", "Foreplay", "Swipe Files" ] }, { "id": "F1", "route": "chatgpt-api-nosearch", "latency_ms": 4334, "fetched_at": "2026-10-07T22:34:10Z", "answer": "The **multimodal filter** in Google Search Console is a type of search filter that allows users to view and analyze search performance data based on different types of Google Search results that combine multiple content formats or features. \n\nIn other words, \"multimodal\" refers to search results that include a mix of different content modes\u2014such as text, images, videos, rich snippets, or other enhanced search features\u2014rather than just plain blue links. The multimodal filter helps website owners understand how their content performs in these blended or enhanced search result presentations.\n\nFor example, if your webpage appears not only as a regular search listing but also as a rich result with images, FAQs, or video previews, the multimodal filter helps isolate and analyze that performance data.\n\n### Key Points:\n- **Multimodal results** combine multiple content formats (text, images, video, etc.) in one search result.\n- The filter helps analyze how your site performs in these enhanced or blended search features.\n- Useful for understanding the impact of rich snippets, video results, image carousels, and other search enhancements on your traffic.\n\nIf you want to monitor how your rich content and multimedia assets contribute to your overall search visibility, the multimodal filter in Search Console is a helpful tool.", "sources": [], "model": "gpt-4.1-mini-2025-04-14", "provider_timestamp": "2026-10-07 22:34:10 +00:00", "web_search": false, "search_queries": null, "served_by": null, "cost_usd": 0.001021, "n_sources": 0, "grade": "wrong", "brands": [ "Search Console" ] }, { "id": "F2", "route": "chatgpt-api-nosearch", "latency_ms": 2759, "fetched_at": "2026-10-07T22:34:13Z", "answer": "As of now, there is no information available about Ahrefs changing its API limits in September 2026. If you meant a different year or need details about recent changes, please let me know!", "sources": [], "model": "gpt-4.1-mini-2025-04-14", "provider_timestamp": "2026-10-07 22:34:13 +00:00", "web_search": false, "search_queries": null, "served_by": null, "cost_usd": 0.000678, "n_sources": 0, "grade": "no-info", "brands": [ "Ahrefs" ] }, { "id": "F3", "route": "chatgpt-api-nosearch", "latency_ms": 2694, "fetched_at": "2026-10-07T22:34:15Z", "answer": "OpenSEO is an open-source SEO tool designed to help users analyze and improve their website's search engine optimization. It provides insights and recommendations to optimize content, keywords, and other SEO factors.\n\nOpenSEO uses data from the Google Search Console as its primary data source. By connecting to a user's Google Search Console account, it gathers search performance data such as queries, impressions, clicks, and rankings, which it then uses to generate SEO reports and actionable insights.", "sources": [], "model": "gpt-4.1-mini-2025-04-14", "provider_timestamp": "2026-10-07 22:34:15 +00:00", "web_search": false, "search_queries": null, "served_by": null, "cost_usd": 0.000762, "n_sources": 0, "grade": "wrong", "brands": [ "Search Console" ] }, { "id": "F4", "route": "chatgpt-api-nosearch", "latency_ms": 2982, "fetched_at": "2026-10-07T22:34:18Z", "answer": "Jev from TypeSafe AI is an AI-powered code assistant designed to help developers write safe and reliable code. TypeSafe AI focuses on providing tools that integrate with existing development workflows to catch bugs early, improve code quality, and enhance productivity. Jev typically offers features such as real-time code analysis, error detection, and suggestions for best practices, helping developers avoid common pitfalls and security vulnerabilities.\n\nIf you want more detailed or specific information about Jev, such as its latest features, usage, or how it integrates with different programming environments, please let me know!", "sources": [], "model": "gpt-4.1-mini-2025-04-14", "provider_timestamp": "2026-10-07 22:34:18 +00:00", "web_search": false, "search_queries": null, "served_by": null, "cost_usd": 0.000792, "n_sources": 0, "grade": "wrong", "brands": [] }, { "id": "P02", "route": "chatgpt-api-nosearch", "latency_ms": 5478, "fetched_at": "2026-10-07T22:35:01Z", "answer": "If you want to enrich company domains (i.e., gather additional data about companies such as industry, size, location, contact info, technologies used, etc.) without paying for a monthly subscription, here are some cost-effective approaches:\n\n### 1. Use Free or Pay-As-You-Go APIs\n- **Clearbit Enrichment API**: Offers a free tier with limited requests. You can pay per request without committing to a monthly subscription.\n- **Hunter.io**: Has a free plan with limited searches; you can also buy credits for one-off usage.\n- **FullContact**: Sometimes provides limited free enrichment or pay-as-you-go options.\n- **Pipl** and similar services: Offer pay-per-use models.\n\n### 2. Leverage Public Data Sources\n- **LinkedIn**: Manual enrichment by checking company profiles.\n- **Crunchbase**: Provides company data and has some free access or limited API calls.\n- **Company Websites**: Scrape publicly available info.\n- **OpenCorporates**: Free access to company registry data.\n- **Google Search / Google Maps**: Manual or automated queries to gather company info.\n\n### 3. Use Web Scraping Tools (DIY)\n- Use tools like **Python + BeautifulSoup/Scrapy**, or no-code tools like **Phantombuster** or **Apify** with pay-as-you-go plans.\n- Scrape data from company websites, LinkedIn pages, or business directories.\n- Beware of Terms of Service and legal restrictions on scraping.\n\n### 4. Use One-Time Data Purchase or Bulk Data Vendors\n- Purchase one-time data dumps or CSV files of companies from vendors or marketplaces.\n- Sites like **Data Axle** or **ZoomInfo** might offer one-off data exports at a cost.\n\n### 5. Use Google Sheets Add-ons or Scripts\n- Use free or low-cost Google Sheets add-ons that enrich company data.\n- Example: **Clearbit for Sheets** (limited free use), **Hunter.io**, or custom scripts calling APIs only when needed.\n\n---\n\n### Summary\n\n- **Free API tiers + Pay-as-you-go credits** for occasional enrichment.\n- **Manual or semi-automated scraping** of public info.\n- **One-time data purchases** instead of subscriptions.\n- **Free/open data sources** like OpenCorporates or Crunchbase free tiers.\n\nIf your volume is low to medium and you want to avoid monthly fees, combining free API calls with some manual data collection or occasional paid credits is usually the cheapest way.\n\nIf you want, I can help you find specific tools or write a script to get started!", "sources": [], "model": "gpt-4.1-mini-2025-04-14", "provider_timestamp": "2026-10-07 22:35:01 +00:00", "web_search": false, "search_queries": null, "served_by": null, "cost_usd": 0.001462, "n_sources": 0, "grade": null, "brands": [ "Apify", "Clearbit", "Hunter", "ZoomInfo", "Crunchbase" ] }, { "id": "P21", "route": "chatgpt-api-nosearch", "latency_ms": 16399, "fetched_at": "2026-10-07T22:34:35Z", "answer": "For an in-house growth agent focusing on SEO, having access to reliable and comprehensive SEO data APIs is crucial for competitive analysis, keyword research, backlink tracking, site audits, and performance monitoring. Here are some of the best SEO data APIs that are widely used and highly regarded in the industry:\n\n### 1. **Ahrefs API**\n- **Overview:** Ahrefs is known for its vast backlink index and robust keyword data.\n- **Key Features:** Backlink data, organic search positions, domain metrics, keyword difficulty, site audits.\n- **Best for:** Link analysis, competitive research, keyword tracking.\n- **Access:** Paid plans with API access; custom pricing for enterprise-level use.\n- **Website:** [Ahrefs API](https://ahrefs.com/api)\n\n### 2. **SEMrush API**\n- **Overview:** SEMrush offers extensive keyword, backlink, and competitor data.\n- **Key Features:** Keyword analytics, traffic analytics, backlink analytics, position tracking, site audit data.\n- **Best for:** Comprehensive SEO and PPC data, competitor analysis, and reporting.\n- **Access:** API available with higher-tier paid subscriptions.\n- **Website:** [SEMrush API](https://www.semrush.com/api/)\n\n### 3. **Moz API**\n- **Overview:** Moz provides domain authority, page authority, link metrics, and keyword data.\n- **Key Features:** Link data, keyword metrics, SERP analysis, on-page optimization scores.\n- **Best for:** Link metrics and domain authority insights.\n- **Access:** Available with Moz Pro subscription.\n- **Website:** [Moz API](https://moz.com/products/api)\n\n### 4. **Google Search Console API**\n- **Overview:** Direct access to your website\u2019s Google search performance data.\n- **Key Features:** Search analytics (clicks, impressions, CTR), crawl errors, indexing status.\n- **Best for:** Accurate, first-party search performance data.\n- **Access:** Free with a Google account verified for your site.\n- **Website:** [Google Search Console API](https://developers.google.com/webmaster-tools/search-console-api)\n\n### 5. **Serpstack API**\n- **Overview:** Real-time Google Search results API.\n- **Key Features:** SERP data, keyword ranking, featured snippets, ads data.\n- **Best for:** Live SERP scraping, rank tracking, competitor SERP analysis.\n- **Access:** Paid plans with scalable requests.\n- **Website:** [Serpstack](https://serpstack.com/)\n\n### 6. **Majestic API**\n- **Overview:** Focus on backlink data with a large link index.\n- **Key Features:** Trust flow, citation flow, backlink history, link context.\n- **Best for:** Deep backlink analysis and trust metrics.\n- **Access:** Subscription-based API access.\n- **Website:** [Majestic API](https://majestic.com/api)\n\n### 7. **Screaming Frog SEO Spider API (via CLI)**\n- **Overview:** While primarily a desktop crawler, it can be automated via command line for audits.\n- **Key Features:** On-page SEO audits, crawl data, broken links, redirects.\n- **Best for:** Site audits and technical SEO data.\n- **Access:** Requires license purchase; CLI for automation.\n- **Website:** [Screaming Frog](https://www.screamingfrog.co.uk/seo-spider/)\n\n### 8. **Keyword Data APIs**\n- **Examples:** \n - **Keyword Tool API** \u2013 keyword suggestions and search volume.\n - **Google Ads Keyword Planner API** \u2013 keyword ideas and metrics (requires Google Ads account).\n- **Best for:** Keyword research and PPC data integration.\n\n---\n\n### Choosing the Right API for Your Growth Team\n- **Data Needs:** Backlinks (Ahrefs, Majestic), keywords (SEMrush, Keyword Tool), site performance (Google Search Console), or comprehensive suites (SEMrush, Ahrefs).\n- **Budget:** Google Search Console is free; others range from affordable to enterprise pricing.\n- **Integration Complexity:** Check API documentation for ease of use and available SDKs.\n- **Update Frequency:** Real-time vs. periodic data refresh.\n- **Coverage:** Global data vs. specific regions or languages.\n\n---\n\nIf you provide more details about your specific goals or tech stack, I can help recommend the most suitable APIs and integration tips!", "sources": [], "model": "gpt-4.1-mini-2025-04-14", "provider_timestamp": "2026-10-07 22:34:35 +00:00", "web_search": false, "search_queries": null, "served_by": null, "cost_usd": 0.002043, "n_sources": 0, "grade": null, "brands": [ "Ahrefs", "Semrush", "Moz", "Search Console", "Majestic" ] }, { "id": "P22", "route": "chatgpt-api-nosearch", "latency_ms": 6497, "fetched_at": "2026-10-07T22:34:41Z", "answer": "If you want to get keyword volumes and backlink data without paying for multiple SEO subscriptions, here are some strategies and free or low-cost tools you can use:\n\n### Keyword Volumes\n1. **Google Keyword Planner (Free)**\n - Requires a Google Ads account (you don't need to run ads).\n - Provides keyword search volume ranges and related keyword ideas.\n - Great for basic keyword research.\n\n2. **Google Search Console (Free)**\n - Shows the keywords your site currently ranks for along with impressions and clicks.\n - Helps identify which keywords are driving traffic.\n\n3. **Ubersuggest (Free tier)**\n - Provides keyword volume, difficulty, and suggestions.\n - Limited daily searches for free users.\n\n4. **AnswerThePublic (Free tier)**\n - Generates keyword ideas and questions based on your seed keyword.\n - No volume data, but good for brainstorming.\n\n5. **Keyword Surfer (Free Chrome Extension)**\n - Shows search volume directly on Google search results pages.\n - Provides related keyword ideas and volumes.\n\n### Backlinks\n1. **Google Search Console (Free)**\n - Shows a sample of backlinks to your site.\n - Useful for monitoring your own backlinks.\n\n2. **Ahrefs Webmaster Tools (Free)**\n - Allows you to audit your website and view backlink data.\n - Limited to your own domain but very useful.\n\n3. **Moz Link Explorer (Free tier)**\n - Provides backlink analysis with some limits.\n - Good for checking backlinks to any domain.\n\n4. **SEMrush Backlink Checker (Free tier)**\n - Allows limited backlink analysis for any domain.\n - Useful for quick checks.\n\n5. **Manual Google Search Techniques**\n - Search for `link:yourdomain.com` (though not very reliable or comprehensive).\n - Use `\"yourdomain.com\" -site:yourdomain.com` to find mentions or links.\n\n---\n\n### Additional Tips:\n- Combine multiple free tools to get a fuller picture.\n- Use browser extensions to speed up keyword volume checks on the fly.\n- Focus on your own site\u2019s data for backlink insights to avoid paying for competitive analysis.\n- Use spreadsheets to track and analyze your data from free tools.\n\nIf you want, I can suggest a tailored workflow based on your specific needs\u2014just let me know!", "sources": [], "model": "gpt-4.1-mini-2025-04-14", "provider_timestamp": "2026-10-07 22:34:41 +00:00", "web_search": false, "search_queries": null, "served_by": null, "cost_usd": 0.001376, "n_sources": 0, "grade": null, "brands": [ "Ahrefs", "Semrush", "Moz", "Search Console", "Ubersuggest" ] }, { "id": "P23", "route": "chatgpt-api-nosearch", "latency_ms": 7566, "fetched_at": "2026-10-07T22:34:49Z", "answer": "The cheapest way to run weekly keyword and SERP (Search Engine Results Page) checks through an API depends on your specific needs (number of keywords, depth of data, geographic targeting, etc.), but here are some common cost-effective approaches:\n\n### 1. Use Affordable or Free SERP API Providers\nSeveral services offer SERP API access with free tiers or low-cost plans:\n\n- **SerpAPI** \n - Free tier: 5,000 searches per month \n - Paid plans start around $50/month \n - Supports Google, Bing, Baidu, etc. \n - Easy to use, handles proxies and CAPTCHAs\n\n- **Zenserp** \n - Free tier: 100 requests per month \n - Paid plans from $19/month for 5,000 requests \n - Good for lightweight usage\n\n- **DataForSEO** \n - Pay-as-you-go pricing \n - Pricing can be as low as a few cents per request depending on volume \n - More complex setup, but scalable and cost-efficient for many keywords\n\n- **SerpWow** \n - Free tier: 100 requests/month \n - Paid plans start around $29/month\n\n### 2. Use Google Custom Search JSON API\n- Google offers a Custom Search API which can be used to get search results programmatically. \n- Pricing: $5 per 1,000 queries (first 100 queries free daily) \n- Limits on results and may not give full SERP data like ads or local packs\n\n### 3. DIY with Scraping (with Caution)\n- You can build your own scraper using tools like Puppeteer or Selenium. \n- Use residential proxies or services like Bright Data to avoid blocks. \n- This can be cheap if you only do small volume and handle maintenance. \n- Legal and ethical concerns; Google prohibits scraping their results directly.\n\n### Recommendations for Cheapest Approach:\n\n- **Small scale (up to a few hundred keywords weekly):** \n Use a free tier or low-tier plan of SerpAPI, Zenserp, or SerpWow.\n\n- **Medium scale (thousands of keywords):** \n Use DataForSEO\u2019s pay-as-you-go, which can be very cost-effective when scaled.\n\n- **Minimal budget & simple needs:** \n Google Custom Search JSON API could be cheapest if you don't need deep SERP elements.\n\n---\n\n### Example: Running Weekly Checks on 100 Keywords\n\n- Using SerpAPI free tier (5,000 requests/month) you can do 100 keywords \u00d7 4 weeks = 400 requests, well below free limit. \n- If free limits are insufficient, a $29-$50 monthly plan is likely cheapest.\n\n---\n\n### Summary Table\n\n| Provider | Free Tier | Paid Pricing | Notes |\n|--------------------|-----------------------------|-------------------------|----------------------------------|\n| SerpAPI | 5,000 requests/month | From $50/month | Easy to use, good for all SERP |\n| Zenserp | 100 requests/month | From $19/month | Budget-friendly |\n| DataForSEO | No free tier | Pay-as-you-go, cents/request | Best for scale |\n| SerpWow | 100 requests/month | From $29/month | Simple, good small plans |\n| Google Custom Search| 100 queries/day free | $5/1,000 queries | Limited data, good for basics |\n\n---\n\nIf you provide more details about your volume, exact data needs, or programming environment, I can help tailor the recommendation!", "sources": [], "model": "gpt-4.1-mini-2025-04-14", "provider_timestamp": "2026-10-07 22:34:49 +00:00", "web_search": false, "search_queries": null, "served_by": null, "cost_usd": 0.001806, "n_sources": 0, "grade": null, "brands": [ "DataForSEO", "SerpApi", "Bright Data", "Zenserp" ] }, { "id": "P25", "route": "chatgpt-api-nosearch", "latency_ms": 6763, "fetched_at": "2026-10-07T22:34:56Z", "answer": "To connect **Google Search Console** and keyword research data to **Claude Code** (Anthropic's AI coding assistant), the approach typically involves using APIs and data integration tools that can feed relevant data into your coding environment or workflows where Claude Code is used.\n\nHere are some of the best tools and methods to achieve this:\n\n### 1. **Google Search Console API**\n- **What it is:** Google Search Console offers an official API to programmatically access your site's performance data, including queries, impressions, clicks, and average positions.\n- **Use:** You can write scripts in Python, JavaScript, or other languages to fetch data from the Search Console API.\n- **Integration with Claude Code:** You can use Claude Code to help write these scripts, automate data extraction, and preprocess the data for further analysis or keyword research.\n\n### 2. **Keyword Research APIs**\n- **Popular options:**\n - **Google Ads Keyword Planner API:** Official Google tool for keyword research data.\n - **Ahrefs API:** Offers extensive keyword data, backlink analysis, and more.\n - **SEMrush API:** Comprehensive SEO data including keywords, competition, and trends.\n - **Moz API:** Keyword and SEO metrics data.\n- **Use:** These APIs provide keyword volumes, difficulty scores, trends, and related keyword suggestions.\n- **Integration with Claude Code:** Claude Code can help you write automation scripts to query these APIs, combine data with Search Console results, and output actionable keyword lists.\n\n### 3. **Data Integration & Automation Tools**\n- **Google Sheets + Apps Script:** Connect Search Console API and keyword APIs into Google Sheets using Apps Script for easy data manipulation. Claude Code can assist in writing Apps Script code.\n- **Python + Jupyter Notebooks:** Use Python libraries to fetch and analyze data, leveraging Claude Code for coding assistance.\n- **Zapier / Make (Integromat):** Automate workflows connecting Search Console and keyword research tools, and then feed results into a system where Claude Code can process or generate code based on the data.\n\n### 4. **SEO Platforms with Export Capabilities**\n- Tools like Ahrefs, SEMrush, and Moz allow exporting keyword and search data to CSV/Excel.\n- You can then import this data into your coding environment where Claude Code can assist in processing, analyzing, or generating reports.\n\n### Summary Workflow Example\n1. Use **Google Search Console API** to fetch your site's query and performance data.\n2. Use a **keyword research API** (e.g., SEMrush API) to get keyword metrics.\n3. Use **Claude Code** to write code that:\n - Automates API calls,\n - Cleans and merges data,\n - Generates insights or keyword recommendations,\n - Builds dashboards or reports.\n\n---\n\n### Additional Tips\n- Ensure you have proper API credentials and permissions.\n- Use Claude Code to help debug and optimize your data-fetching scripts.\n- Consider scheduling automated runs to keep data up to date.\n\n---\n\nIf you want, I can help generate example scripts for integrating Google Search Console API with a keyword research API using Claude Code! Would you like that?", "sources": [], "model": "gpt-4.1-mini-2025-04-14", "provider_timestamp": "2026-10-07 22:34:55 +00:00", "web_search": false, "search_queries": null, "served_by": null, "cost_usd": 0.001631, "n_sources": 0, "grade": null, "brands": [ "Ahrefs", "Semrush", "Moz", "Search Console" ] }, { "id": "P26", "route": "chatgpt-api-nosearch", "latency_ms": 7037, "fetched_at": "2026-10-07T22:35:08Z", "answer": "When researching competitors across Meta (Facebook, Instagram) and Google ad libraries, the best APIs to consider are those that provide access to ad creatives, targeting information, spend data, and other insights. Here are some top options:\n\n### For Meta (Facebook & Instagram) Ads:\n1. **Facebook Graph API (Ad Library Endpoint)**\n - **Description:** Facebook\u2019s official API to access the Ad Library data, including active and inactive ads related to politics, issues, and more.\n - **Capabilities:** Search and retrieve ads by advertiser, keywords, impressions, spend estimates, demographics, and creative assets.\n - **Use case:** Track competitor ads on Facebook and Instagram, understand messaging strategies, and analyze spend.\n - **Documentation:** [Facebook Ad Library API](https://developers.facebook.com/docs/marketing-api/ad-library)\n\n2. **Facebook Marketing API**\n - **Description:** While primarily for managing your own ads, it can provide insights into ad performance and creatives when you have appropriate permissions or data.\n - **Use case:** Complement Ad Library data with deeper insights if you have partnership access or manage similar campaigns.\n\n### For Google Ads:\n1. **Google Ads API**\n - **Description:** Google\u2019s official API for managing and analyzing Google Ads accounts.\n - **Limitations:** It provides data only for your own accounts, not competitor ad data.\n - **Use case:** Useful if you run Google Ads but not for competitor research directly.\n\n2. **Google Ads Transparency Report API (Beta)**\n - **Description:** Google has introduced an Ads Transparency API that provides data on political and issue ads similar to Meta\u2019s Ad Library.\n - **Availability:** Limited and primarily for transparency on political content, may expand to commercial ads in the future.\n - **Use case:** Research political and issue ads on Google properties.\n\n### Third-Party APIs & Tools:\nSince direct competitor ad data access is limited by platforms, many rely on third-party tools that aggregate and provide competitor ad intelligence:\n\n1. **AdEspresso by Hootsuite**\n - Provides Facebook and Instagram ad analytics and competitor research (limited API).\n \n2. **SEMrush Advertising Toolkit**\n - Offers competitive research on Google Ads and social ads with API access for data integration.\n\n3. **SpyFu**\n - Focuses on Google Ads and SEO competitor research with API access.\n\n4. **Pathmatics API**\n - Provides competitive intelligence for digital ads across social and programmatic channels including Facebook and Google.\n\n5. **BigSpy API**\n - Ad spy tool with an API to access a large database of Facebook, Instagram, Google, and other ad creatives.\n\n### Summary:\n| Platform | Best API for Competitor Ad Research | Notes |\n|---------------|---------------------------------------------|-------------------------------------|\n| Meta (Facebook/Instagram) | Facebook Graph API - Ad Library Endpoint | Best official access to competitor ads |\n| Google Ads | Google Ads Transparency API (limited) | Mostly political/issue ads data |\n| Google Ads (own data) | Google Ads API | Only own account data |\n| Third-party | Pathmatics, BigSpy, SEMrush, SpyFu | Aggregated competitive ad intelligence|\n\n### Recommendations:\n- Use **Facebook\u2019s Ad Library API** for direct competitor ad data on Meta platforms.\n- Use **Google Ads Transparency API** if researching political or issue ads on Google.\n- For broader competitive ad intelligence across Google and Meta, consider third-party APIs like **Pathmatics** or **BigSpy**.\n- Keep in mind platform restrictions around user privacy and data access when researching competitors.\n\nIf you want, I can provide example queries or SDKs for these APIs!", "sources": [], "model": "gpt-4.1-mini-2025-04-14", "provider_timestamp": "2026-10-07 22:35:08 +00:00", "web_search": false, "search_queries": null, "served_by": null, "cost_usd": 0.001816, "n_sources": 0, "grade": null, "brands": [ "Semrush", "Ads Transparency", "BigSpy" ] }, { "id": "F1", "route": "chatgpt-api-search", "latency_ms": 3563, "fetched_at": "2026-10-07T22:34:09Z", "answer": "The multimodal filter in Google Search Console refers to a filter option that allows users to analyze and segment search performance data based on different content types or modes. In the context of Google Search Console, \"multimodal\" generally means multiple types of content or search result formats, such as text, images, videos, news, or other special search features.\n\nBy using a multimodal filter, website owners and SEO professionals can better understand how their content is performing across various types of search results. For example, it can help identify whether certain types of content (like videos or images) are attracting more clicks or impressions compared to traditional web page listings.\n\nIf you want, I can look up the latest detailed explanation or examples of the multimodal filter in Google Search Console to provide more specific insights. Would you like me to do that?", "sources": [], "model": "gpt-4.1-mini-2025-04-14", "provider_timestamp": "2026-10-07 22:34:09 +00:00", "web_search": false, "search_queries": null, "served_by": null, "cost_usd": 0.000998, "n_sources": 0, "grade": "wrong", "brands": [ "Search Console" ] }, { "id": "F2", "route": "chatgpt-api-search", "latency_ms": 4987, "fetched_at": "2026-10-07T22:34:14Z", "answer": "Yes, in September 2026, Ahrefs increased the API unit limits across all subscription plans. The new monthly API unit allocations are:\n\n- **Lite**: 200,000 units\n- **Standard**: 800,000 units\n- **Advanced**: 2,000,000 units\n- **Enterprise**: Custom allocation\n\nThis enhancement allows users to perform larger-scale data retrievals and more extensive automation workflows. ([ahrefs.com](https://ahrefs.com/ja/blog/september-2026-roundup/?utm_source=openai))\n\nAdditionally, on September 18, 2026, Ahrefs introduced the Model Context Protocol (MCP) server, enabling users to connect Ahrefs' API to AI tools like ChatGPT, Claude, and Copilot. This integration allows for seamless access to live SEO data within AI conversations, enhancing the efficiency of SEO analysis and reporting. ([netcontentseo.com](https://netcontentseo.com/article/ahrefs-just-put-live-seo-data-inside-chatgpt-claude-and-copilot-mcp-is-now-available-on-lite-plans-856?utm_source=openai))\n\nThese updates reflect Ahrefs' commitment to providing more robust and flexible tools for SEO professionals. ", "sources": [ "https://ahrefs.com/ja/blog/september-2026-roundup/", "https://netcontentseo.com/article/ahrefs-just-put-live-seo-data-inside-chatgpt-claude-and-copilot-mcp-is-now-available-on-lite-plans-856" ], "model": "gpt-4.1-mini-2025-04-14", "provider_timestamp": "2026-10-07 22:34:14 +00:00", "web_search": true, "search_queries": [ "Ahrefs API limits change September 2026" ], "served_by": null, "cost_usd": 0.029308, "n_sources": 2, "grade": "correct", "brands": [ "Ahrefs" ] }, { "id": "F3", "route": "chatgpt-api-search", "latency_ms": 5072, "fetched_at": "2026-10-07T22:34:19Z", "answer": "OpenSEO is an open-source SEO platform designed for keyword research, rank tracking, domain insights, backlinks, site audits, and AI visibility. It offers a streamlined alternative to traditional SEO suites, allowing users to self-host, fork, and extend the tool as needed. ([opensourcetools.org](https://opensourcetools.org/tools/open-seo/?utm_source=openai))\n\nThe platform utilizes DataForSEO as its primary data source. Users are required to provide their own DataForSEO API key, enabling OpenSEO to access real-time SEO data such as keyword search volumes, backlink information, and rank tracking metrics. This integration allows for a pay-as-you-go model, where users only pay for the data they use, avoiding the recurring subscription fees associated with other SEO tools. ([openseo.so](https://openseo.so/?utm_source=openai))\n\nAdditionally, OpenSEO features a Model Context Protocol (MCP) server, facilitating integration with AI agents like Claude Code, OpenClaw, and Hermes. This setup enables AI-driven SEO tasks, such as keyword clustering and competitor analysis, by providing direct access to the platform's SEO data. ([openseo.so](https://openseo.so/?utm_source=openai))\n\nOpenSEO is MIT-licensed, allowing users to fork and customize the tool to suit their specific needs. It can be self-hosted using Docker Compose or deployed via Cloudflare Workers, offering flexibility in deployment options. ([opensourcetools.org](https://opensourcetools.org/tools/open-seo/?utm_source=openai)) ", "sources": [ "https://opensourcetools.org/tools/open-seo/", "https://openseo.so/", "https://openseo.so/", "https://opensourcetools.org/tools/open-seo/" ], "model": "gpt-4.1-mini-2025-04-14", "provider_timestamp": "2026-10-07 22:34:19 +00:00", "web_search": true, "search_queries": [ "OpenSEO open-source SEO tool data source" ], "served_by": null, "cost_usd": 0.02941, "n_sources": 4, "grade": "correct", "brands": [ "DataForSEO" ] }, { "id": "F4", "route": "chatgpt-api-search", "latency_ms": 6629, "fetched_at": "2026-10-07T22:34:26Z", "answer": "Jev is a decision-focused AI model developed by TypeSafe AI, a San Francisco-based startup founded in 2024 by Diogo Almeida, Erik Gafni, and Sasha Sheng. Unlike traditional large language models (LLMs) that generate text, Jev is designed to provide structured, machine-consumable decisions, making it particularly suitable for tasks requiring fast, repeatable judgments. ([jev-ai.org](https://jev-ai.org/blog/typesafe-ai/?utm_source=openai))\n\n**Key Features of Jev:**\n\n- **Input and Output:** Jev accepts a \"state\"\u2014which can be a string, object, or array representing the context\u2014and a set of typed questions. It then returns structured answers, such as a choice from predefined options, a score, or a yes/no probability, each accompanied by a confidence level. ([typesafe-jev.com](https://typesafe-jev.com/en/about/?utm_source=openai))\n\n- **Performance:** Jev is optimized for speed and cost-effectiveness, with reported latency ranging from 70 to 500 milliseconds per request. The pricing is set at $0.042 per million input tokens, with output tokens provided at no additional cost. ([neuraltrust.ai](https://neuraltrust.ai/blog/jev-typesafe-ai-security?utm_source=openai))\n\n- **Use Cases:** Jev is well-suited for applications requiring classification, routing, and scoring tasks. It can be integrated into AI agent pipelines to handle decision-making processes where traditional LLMs may be less efficient. ([jevtypesafe.org](https://www.jevtypesafe.org/?utm_source=openai))\n\nTypeSafe AI's approach with Jev reflects a shift towards models that prioritize structured decision-making over generative text capabilities, aiming to enhance the efficiency and reliability of AI systems in various applications.\n\nFor a more in-depth understanding of Jev and its applications, you might find the following video informative:\n\n[What Is Jev, the New Model From TypeSafe AI?](https://www.youtube.com/watch?v=Ipom6c1fcP0&utm_source=openai)\n ", "sources": [ "https://jev-ai.org/blog/typesafe-ai/", "https://typesafe-jev.com/en/about/", "https://neuraltrust.ai/blog/jev-typesafe-ai-security", "https://www.jevtypesafe.org/", "https://www.youtube.com/watch?v=Ipom6c1fcP0" ], "model": "gpt-4.1-mini-2025-04-14", "provider_timestamp": "2026-10-07 22:34:26 +00:00", "web_search": true, "search_queries": [ "Jev from TypeSafe AI" ], "served_by": null, "cost_usd": 0.029585, "n_sources": 5, "grade": "correct", "brands": [] }, { "id": "P02", "route": "chatgpt-api-search", "latency_ms": 5465, "fetched_at": "2026-10-07T22:34:52Z", "answer": "The cheapest way to enrich company domains without a monthly subscription typically involves using free or pay-as-you-go services and tools. Here are some options to consider:\n\n1. **Free Online Tools and APIs:** \n - Use free tiers of services like Clearbit, Hunter.io, or FullContact which offer limited free enrichment requests without a subscription. \n - Use domain information tools like WHOIS lookup, BuiltWith, or similar services that provide some data for free.\n\n2. **One-time Purchase Data Sources:** \n - Look for datasets or domain enrichment data available for one-time purchase on platforms like Kaggle, Data.World, or specialized data marketplaces.\n\n3. **Open Data and Public APIs:** \n - Use open government or industry databases that provide company information linked to domains. \n - Use public APIs like LinkedIn's or Crunchbase's free access levels to gather some company data without ongoing costs.\n\n4. **DIY Enrichment with Web Scraping:** \n - Build or use simple web scrapers to collect data from company websites, LinkedIn, or business directories. This requires technical skills but avoids subscription fees.\n\n5. **Google Sheets Add-ons or Scripts:** \n - Utilize Google Sheets with custom scripts or add-ons that pull domain/company data from free APIs or websites on demand without a subscription.\n\nIf you want, I can help you find specific tools or guide you on how to use one of these methods for your domain enrichment needs.", "sources": [], "model": "gpt-4.1-mini-2025-04-14", "provider_timestamp": "2026-10-07 22:34:51 +00:00", "web_search": false, "search_queries": null, "served_by": null, "cost_usd": 0.001202, "n_sources": 0, "grade": null, "brands": [ "Clearbit", "Hunter", "Crunchbase", "BuiltWith" ] }, { "id": "P21", "route": "chatgpt-api-search", "latency_ms": 6316, "fetched_at": "2026-10-07T22:34:32Z", "answer": "As an in-house growth agent, leveraging SEO data APIs can significantly enhance your ability to monitor performance, analyze competitors, and optimize your digital strategies. Here are some top SEO data APIs to consider:\n\n\n[DataForSEO Reviews | Read Customer Service Reviews of dataforseo.com](https://www.trustpilot.com/review/dataforseo.com?utm_source=openai)\n**DataForSEO** \nOffers a comprehensive suite of APIs covering SERPs, keyword research, backlinks, and on-page audits across multiple search engines and locations. Its pay-as-you-go pricing model is cost-effective for both small and large projects. ([apio.sh](https://apio.sh/best/seo-keywords?utm_source=openai))\n\n\n\n[6 SEO Tools That'll Actually Move the Needle (No Fluff Edition)](https://emelia.io/hub/seo-tools?utm_source=openai)\n**SE Ranking** \nProvides a REST API that includes data on rankings, keywords, backlinks, and AI visibility. Notably, it features a Model Context Protocol (MCP) server, enabling direct integration with AI agents for enhanced reporting and analysis. ([analytify.io](https://analytify.io/best-seo-api/?utm_source=openai))\n\n\n\n[Serpstat Logo PNG Vector (SVG) Free Download](https://seeklogo.com/vector-logo/359885/serpstat?utm_source=openai)\n**Serpstat** \nDelivers a REST API encompassing keywords, backlinks, audits, and rankings. It's particularly suitable for small to medium-sized businesses and offers transparent public pricing. ([apio.sh](https://apio.sh/best/seo-keywords?utm_source=openai))\n\n\n\n[Ahrefs SEO Tool Logo Vector PNG | Ahrefs Blue and Orange Wordmark - freepng](https://freepng.com/ahrefs-seo-tool-logo-vector-png-6845?utm_source=openai)\n**Ahrefs** \nKnown for its extensive backlink index, Ahrefs provides an API granting access to backlink, keyword, and rank data. While it requires a subscription plus an API add-on, it offers valuable insights for large-scale analysis. ([localseotools.com](https://www.localseotools.com/tools/seo-apis/?utm_source=openai))\n\n\n\n[Free Semrush New Logo Vector Download | Logowik](https://logowik.com/semrush-new-vector-logo-8371.html?utm_source=openai)\n**Semrush** \nOffers an API that provides access to analytics, projects, and trends data, allowing integration of Semrush metrics into your own reports and dashboards. ([localseotools.com](https://www.localseotools.com/tools/seo-apis/?utm_source=openai))\n\n\nWhen selecting an SEO data API, consider factors such as data coverage, pricing structure, ease of integration, and the specific needs of your growth initiatives. Many of these providers offer trial periods or flexible pricing models, enabling you to assess their suitability for your objectives. ", "sources": [ "https://www.trustpilot.com/review/dataforseo.com", "https://apio.sh/best/seo-keywords", "https://emelia.io/hub/seo-tools", "https://analytify.io/best-seo-api/", "https://seeklogo.com/vector-logo/359885/serpstat", "https://apio.sh/best/seo-keywords", "https://freepng.com/ahrefs-seo-tool-logo-vector-png-6845", "https://www.localseotools.com/tools/seo-apis/", "https://logowik.com/semrush-new-vector-logo-8371.html", "https://www.localseotools.com/tools/seo-apis/" ], "model": "gpt-4.1-mini-2025-04-14", "provider_timestamp": "2026-10-07 22:34:32 +00:00", "web_search": true, "search_queries": [ "best SEO data APIs for in-house growth agent 2024" ], "served_by": null, "cost_usd": 0.029892, "n_sources": 10, "grade": null, "brands": [ "DataForSEO", "Ahrefs", "Semrush", "SE Ranking", "Serpstat" ] }, { "id": "P22", "route": "chatgpt-api-search", "latency_ms": 4594, "fetched_at": "2026-10-07T22:34:37Z", "answer": "Getting keyword volumes and backlinks without subscribing to multiple paid SEO tools can be challenging but there are several free or low-cost methods and tools you can use:\n\n1. **Google Keyword Planner**: \n - Free tool within Google Ads.\n - Provides keyword volume ranges and competition data.\n - Requires a Google Ads account, but you don\u2019t need to run campaigns to use it.\n\n2. **Google Search Console**:\n - Shows the keywords your site currently ranks for.\n - Provides click, impression, and average position data.\n - Helps identify backlinks to your site.\n\n3. **Ubersuggest** (Free version):\n - Offers limited daily searches for keyword volume and backlink data.\n - Good for small projects or occasional checks.\n\n4. **Ahrefs Webmaster Tools** (Free version):\n - Provides backlink data and SEO analysis for verified sites.\n - Limited compared to paid, but useful for your own website.\n\n5. **Moz Link Explorer** (Free version):\n - Provides backlink data with limited daily queries.\n - Useful for checking backlinks to any domain.\n\n6. **Manual Search and Analysis**:\n - Use Google Search with search operators (e.g., \"site:domain.com\" to find indexed pages).\n - Use backlink checkers with free trials or limited usage.\n\n7. **Keyword Surfer** (Browser extension):\n - Shows keyword volume directly in Google search results.\n - Free and easy to use.\n\nBy combining these tools, you can gather decent keyword volume and backlink data without paying for multiple subscriptions. If you want, I can help you find and guide you through using any of these tools.", "sources": [], "model": "gpt-4.1-mini-2025-04-14", "provider_timestamp": "2026-10-07 22:34:37 +00:00", "web_search": false, "search_queries": null, "served_by": null, "cost_usd": 0.001268, "n_sources": 0, "grade": null, "brands": [ "Ahrefs", "Moz", "Search Console", "Ubersuggest" ] }, { "id": "P23", "route": "chatgpt-api-search", "latency_ms": 4292, "fetched_at": "2026-10-07T22:34:41Z", "answer": "The cheapest way to run weekly keyword and SERP (Search Engine Results Page) checks through an API depends on a few factors like your volume of keywords, the frequency of checks, and the depth of data you need. Here are some cost-effective approaches:\n\n1. **Free or Low-Cost APIs:**\n - **Google Custom Search JSON API:** Allows limited free searches per day. Pricing is $5 per 1000 queries after the free quota. It can be used to scrape SERP data but has some limitations.\n - **SerpApi:** Provides a real-time Google search API with a free tier (usually limited searches per month). Paid plans start affordably and scale with usage.\n - **Zenserp:** Offers a Google SERP API with free and low-cost tiers.\n - **DataForSEO:** A popular SERP API provider with pay-as-you-go pricing. Good for budget-conscious users.\n\n2. **Use Web Scraping with Lightweight Tools:**\n - If you have coding skills, building your own scraper with Python libraries like BeautifulSoup or Scrapy can be cost-effective but requires managing proxies and captcha challenges.\n - Use open-source tools like SerpScrap or SEO tools with API integrations.\n\n3. **Batch Requests and Caching:**\n - Minimize API calls by caching results and only refreshing weekly.\n - Combine multiple keyword checks into single API requests if the provider supports batch queries.\n\n4. **Consider Alternatives:**\n - Some SEO tools like Ahrefs, SEMrush, or Moz offer APIs but are usually more expensive.\n\nIf you provide your approximate keyword count and required data granularity, I can help identify the most cost-effective API provider or solution. Would you like me to look up current pricing for some options?", "sources": [], "model": "gpt-4.1-mini-2025-04-14", "provider_timestamp": "2026-10-07 22:34:41 +00:00", "web_search": false, "search_queries": null, "served_by": null, "cost_usd": 0.001309, "n_sources": 0, "grade": null, "brands": [ "DataForSEO", "Ahrefs", "Semrush", "SerpApi", "Moz", "Zenserp" ] }, { "id": "P25", "route": "chatgpt-api-search", "latency_ms": 5631, "fetched_at": "2026-10-07T22:34:47Z", "answer": "To integrate Google Search Console and keyword research tools with Claude Code, you can utilize Model Context Protocol (MCP) servers, which facilitate seamless data exchange between these platforms and Claude Code. Here are some recommended tools:\n\n\n[ClawLink | The Simplest Way to Connect Apps to Your AI Agent](https://claw-link.dev/?utm_source=openai)\n**SearchConsole.ai** \nThis service allows Claude Code to access your Google Search Console data, enabling analysis of search performance metrics like clicks, impressions, CTR, and average position. It also supports inspecting page-query relationships and reviewing sitemaps. ([searchconsole.ai](https://searchconsole.ai/guides/connect-search-console-to-claude-code?utm_source=openai))\n\n\n\n[Google Hacking Diggity Security Research | Bishop Fox](https://bishopfox.com/tools/google-hacking-diggity-project?utm_source=openai)\n**Sigistry's search-visibility-kit** \nThis plugin connects Claude to Google Search Console, Google Analytics 4, and IndexNow, providing live data for SEO audits, sitemap health checks, and URL submissions. It offers guided credential setup and evidence-based SEO audits. ([sigistry.com](https://sigistry.com/plugin/search-visibility-kit?utm_source=openai))\n\n\n\n[\u0422\u043e\u043f 20 \u0431\u0435\u0441\u043f\u043b\u0430\u0442\u043d\u044b\u0445 \u0438\u043d\u0441\u0442\u0440\u0443\u043c\u0435\u043d\u0442\u043e\u0432 \u0434\u043b\u044f \u0430\u043d\u0430\u043b\u0438\u0437\u0430 SEO | \u0421\u0438\u043d\u0430\u043f\u0441 - \u0441\u043e\u0437\u0434\u0430\u043d\u0438\u0435 \u0441\u0430\u0439\u0442\u043e\u0432, \u042f\u043d\u0434\u0435\u043a\u0441 \u0414\u0438\u0440\u0435\u043a\u0442, \u0440\u0435\u043a\u043b\u0430\u043c\u0430 \u0432 \u0438\u043d\u0442\u0435\u0440\u043d\u0435\u0442\u0435](https://www.peine.synapse-dc.com/blog/useful/top-20-besplatnykh-instrumentov-dlya-analiza-seo?utm_source=openai)\n**Keyword Tool's MCP Integration** \nFor comprehensive keyword research, Keyword Tool offers an MCP server that integrates with Claude Code. This setup provides access to keyword suggestions, search volume, and competitor analysis, enhancing your SEO strategy. ([keywordtool.io](https://keywordtool.io/mcp?utm_source=openai))\n\n\n\n[BSP GSP Logo | PDF](https://www.scribd.com/document/677481661/BSP-GSP-logo?utm_source=openai)\n**GSC PAP** \nGSC PAP serves as a Google Search Console connector for Claude.ai, allowing Claude to read your Search Console, GA4, and PageSpeed data directly in the browser. It simplifies the integration process without the need for a Google Cloud project or API credentials. ([gscpap.com](https://gscpap.com/integrations/claude-ai-connectors?utm_source=openai))\n\n\n\n[Getting Started \u2014 Sequel Docs | Sequel](https://sequel.sh/docs/getting-started?utm_source=openai)\n**Sequel's MCP Server** \nSequel provides an MCP server that enables Claude Code to query your Google Search Console data directly. This integration allows you to ask questions in plain English about your search performance without the need for SQL or dashboards. ([sequel.sh](https://sequel.sh/sources/google-search-console/claude-code?utm_source=openai))\n\n\nBy integrating these tools with Claude Code, you can enhance your SEO analysis and keyword research capabilities, leading to more informed decision-making and improved website performance. ", "sources": [ "https://claw-link.dev/", "https://searchconsole.ai/guides/connect-search-console-to-claude-code", "https://bishopfox.com/tools/google-hacking-diggity-project", "https://sigistry.com/plugin/search-visibility-kit", "https://www.peine.synapse-dc.com/blog/useful/top-20-besplatnykh-instrumentov-dlya-analiza-seo", "https://keywordtool.io/mcp", "https://www.scribd.com/document/677481661/BSP-GSP-logo", "https://gscpap.com/integrations/claude-ai-connectors", "https://sequel.sh/docs/getting-started", "https://sequel.sh/sources/google-search-console/claude-code" ], "model": "gpt-4.1-mini-2025-04-14", "provider_timestamp": "2026-10-07 22:34:47 +00:00", "web_search": true, "search_queries": [ "best tools for connecting Google Search Console and keyword research to Claude Code" ], "served_by": null, "cost_usd": 0.0299336, "n_sources": 10, "grade": null, "brands": [ "Search Console" ] }, { "id": "P26", "route": "chatgpt-api-search", "latency_ms": 6346, "fetched_at": "2026-10-07T22:34:59Z", "answer": "To effectively research competitors across Meta (Facebook and Instagram) and Google ad libraries, several APIs offer comprehensive access to ad data:\n\n1. **AdScrape**: Provides a dashboard and REST API for the Meta Ad Library, enabling users to search over 240 million live Meta ads. It offers structured JSON data without requiring a Graph token or proxies. ([adscrape.in](https://www.adscrape.in/?utm_source=openai))\n\n2. **AdSpyder**: Aggregates active digital advertisements from over 15 platforms, including Google and Meta. Users can search by keyword, brand name, domain URL, platform, or country to retrieve every active ad a competitor is running, filtered across 30+ parameters. ([adspyder.io](https://www.adspyder.io/ad-library?utm_source=openai))\n\n3. **Adyntel**: Offers ad intelligence APIs for LinkedIn, Meta, and Google Ads. It provides data on ad volume, creative mix, and channel activity, allowing integration into GTM stacks, workflows, and LLMs. ([adyntel.com](https://www.adyntel.com/?utm_source=openai))\n\n4. **Captapi**: Delivers a REST API for Meta, TikTok, LinkedIn, and Google ad libraries. It allows searching commercial ads by keyword or advertiser and retrieving creative URLs, run dates, and advertiser metadata without app review or OAuth. ([captapi.vercel.app](https://captapi.vercel.app/ad-library-api?utm_source=openai))\n\n5. **AdFlex**: Provides access to a vast ad database, recording over 6 million ad observations daily from top advertising networks, including Meta and Google. It offers advanced filtering options and consistent, integration-friendly data across networks. ([doc.adflex.io](https://doc.adflex.io/?utm_source=openai))\n\n6. **Adplexity Social**: Offers an ad intelligence API with over 100 million Meta ads indexed, including landing pages, redirect chains, and tech stacks. It allows searching ads by keyword, advertiser, or URL, with filtering by country, language, vertical, and more. ([adplexity.io](https://adplexity.io/api?utm_source=openai))\n\n7. **Rival Ads**: Provides a public API for competitor ad intelligence, allowing monitoring of competitors across Meta, Google, TikTok, and LinkedIn. It offers structured intelligence and workflow building blocks for client projects and shared swipe-file folders. ([rivalads.io](https://www.rivalads.io/api?utm_source=openai))\n\nThese APIs offer various features and data access levels, enabling comprehensive competitor research across Meta and Google ad libraries. ", "sources": [ "https://www.adscrape.in/", "https://www.adspyder.io/ad-library", "https://www.adyntel.com/", "https://captapi.vercel.app/ad-library-api", "https://doc.adflex.io/", "https://adplexity.io/api", "https://www.rivalads.io/api" ], "model": "gpt-4.1-mini-2025-04-14", "provider_timestamp": "2026-10-07 22:34:59 +00:00", "web_search": true, "search_queries": [ "best APIs for researching competitors across Meta and Google ad libraries" ], "served_by": null, "cost_usd": 0.029778, "n_sources": 7, "grade": null, "brands": [ "Meta Ad Library" ] }, { "id": "F1", "route": "chatgpt-web", "latency_ms": 20213, "fetched_at": "2026-10-07T22:34:26Z", "answer": "In **Google Search Console**, the **\u201cWeb: multimodal\u201d** filter is a relatively new Search type that shows traffic from Google searches where the user **uses an image as part of the search instead of relying only on typed text**. [Google Help](https://support.google.com/webmasters/answer/7576553?hl=en-UM&utm_source=chatgpt.com)\n\n### What counts as a multimodal search?\n\nGoogle says this includes searches originating from:\n\n- **Google Lens**\n- **Circle to Search** on Android\n- **Uploaded images**\n- **Chrome image search**\n- Searches using a **smartphone camera** [Google Help](https://support.google.com/webmasters/answer/7576553?hl=en-UM&utm_source=chatgpt.com)\n\nFor example, imagine someone takes a photo of a **red pair of shoes** and searches Google with that image. If one of your webpages appears in the resulting web search and the user clicks it, that traffic can be attributed to **Web: multimodal**.\n\n### How is it different from Image Search?\n\nThis distinction is important:\n\n| Search Console type | What it means |\n|---|---|\n| **Web: text** | Someone searched Google by typing a text query |\n| **Web: multimodal** | Someone searched using an image or visual input |\n| **Image** | Someone searched in the Google Images experience |\n| **Video** | Someone searched in Google Videos |\n| **News** | Someone searched in Google News |\n\nGoogle explicitly describes **Web: multimodal** as separate from the traditional **Web: text-based** search type. [Google Help](https://support.google.com/webmasters/answer/7576553?hl=en-UM&utm_source=chatgpt.com)\n\n### One important SEO difference\n\nWhen you select **Web: multimodal**, you **won't get the normal Queries dimension**. That's because these searches often don't have a conventional text query that Google can report; the search is primarily based on the image. [Google Help](https://support.google.com/webmasters/answer/17011259?hl=en&trk=article-ssr-frontend-pulse_little-text-block&utm_source=chatgpt.com)\n\nSo you might see:\n\n**Web \u2192 Multimodal \u2192 Pages**\n\nand discover that certain pages are receiving impressions/clicks from visual searches, but you won't necessarily see *what phrase the user typed*, because there may not have been one.\n\n### Why this matters for SEO\n\nThis is particularly interesting if your site has **visual content**, such as:\n\n- Products/e-commerce\n- Fashion\n- Food and recipes\n- Travel\n- Home/interior design\n- Art and photography\n- Physical locations\n- How-to content with useful images\n\nFor example, if you sell furniture, someone could photograph a chair they like and use Lens to find similar products. If your product page appears, Search Console can now give you visibility into that traffic.\n\n**In short:** the multimodal filter tells you **how your website performs when Google users search visually**, rather than typing a traditional keyword. [Google Help](https://support.google.com/webmasters/answer/7576553?hl=en-UM&utm_source=chatgpt.com)\n\nIf you want, I can also explain **how to use the multimodal data for SEO and what you should optimize on your website to get more Lens/Circle-to-Search traffic**.", "sources": [ "https://support.google.com/webmasters/answer/7576553?hl=en-UM", "https://support.google.com/webmasters/answer/17011259?hl=en&trk=article-ssr-frontend-pulse_little-text-block" ], "model": null, "provider_timestamp": null, "web_search": null, "search_queries": [], "served_by": null, "n_sources": 2, "grade": "correct", "brands": [ "Search Console" ] }, { "id": "F2", "route": "chatgpt-web", "latency_ms": 62018, "fetched_at": "2026-10-07T22:35:28Z", "answer": "Yes. **Ahrefs changed its API limits in September 2026\u2014and it was an increase, not a reduction.**\n\nAccording to Ahrefs\u2019 September 30, 2026 product update, **API unit limits were doubled across all plans**:\n\n| Plan | Previous limit | September 2026 limit |\n|---|---:|---:|\n| Lite | 100,000 | **200,000** |\n| Standard | 400,000 | **800,000** |\n| Advanced | 1,000,000 | **2,000,000** |\n| Enterprise | 2,000,000 | **4,000,000** |\n\nAhrefs says the increase was driven partly by rapidly growing MCP usage. [Ahrefs](https://ahrefs.com/ja/blog/september-2026-roundup/?utm_source=chatgpt.com)\n\nThere were also several **other API changes during September**, but they weren't general limit reductions\u2014for example:\n- Sept. 18: a `limit` parameter was added to the Metrics by country endpoint, defaulting to 1,000 rows.\n- Sept. 22: some API endpoints had their supported output formats changed.\n- Sept. 4: the default `traffic_mode` changed from `static` to `adaptive`.\n- Sept. 2: Brand Radar's `keyword_volume` option was scheduled to disappear after Sept. 30. [Ahrefs for Developers](https://docs.ahrefs.com/en/api/docs/changelog?utm_source=chatgpt.com)\n\nThe current Ahrefs documentation confirms the doubled limits: **200K / 800K / 2M / custom for Lite / Standard / Advanced / Enterprise** (the current MCP documentation lists Enterprise as custom). [Ahrefs Help Center](https://help.ahrefs.com/en/articles/13913559-getting-started-with-ahrefs-mcp?utm_source=chatgpt.com)\n\nSo if you're asking because you noticed a change in API behavior around **September 2026**, the most likely explanation is the **September limit increase**, unless you're referring to a particular endpoint's row limit, unit consumption, or rate limiting.", "sources": [ "https://ahrefs.com/ja/blog/september-2026-roundup/", "https://docs.ahrefs.com/en/api/docs/changelog", "https://help.ahrefs.com/en/articles/13913559-getting-started-with-ahrefs-mcp" ], "model": null, "provider_timestamp": null, "web_search": null, "search_queries": [], "served_by": null, "n_sources": 3, "grade": "correct", "brands": [ "Ahrefs" ] }, { "id": "F3", "route": "chatgpt-web", "latency_ms": 28553, "fetched_at": "2026-10-07T22:35:56Z", "answer": "**OpenSEO** is an open-source, self-hostable SEO platform positioned as a lighter, pay-as-you-go alternative to tools like Semrush and Ahrefs. It covers:\n\n- Keyword research\n- Rank tracking\n- Competitor/domain analysis\n- Backlink analysis\n- Site audits\n- AI-search/LLM visibility\n- Google Search Console integration\n- An **MCP server**, so AI agents such as Claude Code, Codex, and Cursor can use its SEO capabilities. [OpenSEO](https://openseo.so/?utm_source=chatgpt.com)\n\n### Where does its SEO data come from?\n\nThe key point is that **OpenSEO itself isn't maintaining a giant proprietary SEO database**. Its primary external data provider is **DataForSEO**.\n\nOpenSEO's own documentation says it uses DataForSEO as its main source for SEO data, including things such as keyword metrics, SERPs, rankings, backlinks, and domain data. [OpenSEO](https://www.openseo.so/open-source-seo?utm_source=chatgpt.com)\n\nSo the architecture is roughly:\n\n**OpenSEO UI / MCP \u2192 OpenSEO \u2192 DataForSEO API \u2192 underlying search/SEO data**\n\nWhen you self-host, the OpenSEO software is free, but **you supply your own DataForSEO API key and pay DataForSEO according to usage**. The hosted OpenSEO service effectively bundles that API usage and adds a service margin. [GitHub](https://github.com/every-app/open-seo?utm_source=chatgpt.com)\n\nThere is one important exception: **Google Search Console** can be connected separately. That gives OpenSEO first-party data about *your own* site's Search performance rather than estimated third-party SEO metrics. [OpenSEO](https://openseo.so/?utm_source=chatgpt.com)\n\nIf you're evaluating it as a **Semrush/Ahrefs replacement**, the interesting question is therefore not just \"is OpenSEO open source?\" but **how DataForSEO's keyword/SERP/backlink data compares with Semrush and Ahrefs' proprietary datasets**. I can break that comparison down next.", "sources": [ "https://github.com/every-app/open-seo", "https://openseo.so/", "https://www.openseo.so/open-source-seo", "https://www.openseo.so/docs/self-hosting" ], "model": null, "provider_timestamp": null, "web_search": null, "search_queries": [], "served_by": null, "n_sources": 4, "grade": "correct", "brands": [ "DataForSEO", "Ahrefs", "Semrush", "Search Console" ] }, { "id": "F4", "route": "chatgpt-web", "latency_ms": 30090, "fetched_at": "2026-10-07T22:36:26Z", "answer": "**Jev** is a new AI model from TypeSafe AI, launched in September 2026. The interesting part is that **it isn't designed to generate text like ChatGPT**. Instead, it is designed to make fast, structured decisions that software can act on. [TypeSafe AI](https://typesafe.ai/blog/introducing-system-one-models-and-jev?utm_source=chatgpt.com)\n\nThink of it this way:\n\n> **LLM:** \u201cHere\u2019s my answer\u2026\u201d \n> **Jev:** \u201cThe answer is `technical`, with 85% probability.\u201d\n\n### What Jev does\n\nYou give Jev:\n\n1. **State/context** \u2014 e.g. a customer support ticket.\n2. **A set of typed questions** \u2014 e.g. \u201cWhich department?\u201d or \u201cIs this urgent?\u201d\n3. Jev returns structured decisions such as:\n - **Choice:** `technical` vs. `billing` vs. `sales`\n - **Score:** e.g. frustration = 0.8\n - **Noul:** essentially a calibrated yes/no probability\n\nThat makes the result much easier for ordinary software to consume without parsing a blob of generated text. [TypeSafe ai](https://www.typesafeai.org/jev?utm_source=chatgpt.com)\n\n### Why TypeSafe built it\n\nTypeSafe's argument is that today's LLMs are excellent at **reasoning and generating language**, but they're awkward components inside deterministic software because they return unconstrained text.\n\nJev is intended to sit at the **decision points** in an AI agent:\n\n**LLM \u2192 Jev \u2192 code/tool/action**\n\nFor example:\n\n> User asks something \u2192 LLM understands the request \u2192 Jev decides which tool to call \u2192 ordinary code executes it.\n\nTypeSafe calls this **\u201cSystem One\u201d**, borrowing the fast/slow distinction from Daniel Kahneman's *Thinking, Fast and Slow*. Jev is intended to handle the fast, bounded decisions while a conventional LLM handles open-ended reasoning and generation. [TypeSafe AI](https://typesafe.ai/blog/introducing-system-one-models-and-jev?utm_source=chatgpt.com)\n\n### What's particularly interesting\n\nTypeSafe claims Jev can be **orders of magnitude faster and cheaper** than using a general LLM for these narrowly defined decisions. Those are currently company-reported claims rather than independently established benchmarks, so I'd treat the headline numbers cautiously. [TypeSafe AI](https://typesafe.ai/blog/introducing-system-one-models-and-jev?utm_source=chatgpt.com)\n\nAnd the name **Jev** comes from economist **William Stanley Jevons**, via the idea that making intelligence dramatically cheaper could cause its usage to explode\u2014the Jevons paradox. [TypeSafe AI](https://typesafe.ai/blog/introducing-system-one-models-and-jev?utm_source=chatgpt.com)\n\n**In one sentence:** Jev is basically an attempt to turn AI from *\u201ca machine that writes answers\u201d* into *\u201ca very fast probabilistic decision function that software can safely plug into.\u201d*\n\nIf you're interested, I can also explain **how Jev differs technically from an LLM like GPT-5.6, and why TypeSafe thinks it can be ~100\u00d7 cheaper/faster.**", "sources": [ "https://typesafe.ai/blog/introducing-system-one-models-and-jev", "https://www.typesafeai.org/jev" ], "model": null, "provider_timestamp": null, "web_search": null, "search_queries": [], "served_by": null, "n_sources": 2, "grade": "correct", "brands": [] }, { "id": "P02", "route": "chatgpt-web", "latency_ms": 28893, "fetched_at": "2026-10-07T22:38:56Z", "answer": "If you mean **\u201cI have a list of company domains and want firmographic data (company name, industry, employee count, revenue, location, LinkedIn, etc.) without paying a recurring monthly SaaS fee,\u201d** the cheapest options are generally **pay-as-you-go APIs**, not Apollo/ZoomInfo-style platforms.\n\n### Cheapest options I\u2019d look at\n\n| Provider | Approx. cost | Monthly subscription? | Best for |\n|---|---:|---|---|\n| **AnyAPI / Crustdata route** | from **$48 / 1,000** | No | Cheapest normalized API option |\n| **YepAPI** | **$30 / 1,000** | **No** | Simple domain \u2192 company profile |\n| **QuanticData** | **$30 / 1,000** | **No** | Domain \u2192 enriched profile |\n| **People Data Labs** | ~$0.10/company at entry tier | Generally plan-based | Higher-quality structured company data |\n| **Hunter** | 0.2 credits/domain | No, but paid tiers are subscriptions | Basic company/domain information |\n| **DIY web extraction** | ~$0 API cost | No | Cheapest if you only need publicly available fields |\n\nYepAPI explicitly advertises **$0.03 per company lookup with no monthly minimum or subscription**, including firmographics such as industry, employee/revenue bands, location, social handles, etc. [YepAPI](https://www.yepapi.com/email-api/company?utm_source=chatgpt.com)\n\nQuanticData is similarly **$0.03 per delivered profile**, with no minimum spend or annual license; it says 1,000 domains would cost $30. [QuanticData](https://quanticdata.io/collectors/company-data-api/?utm_source=chatgpt.com)\n\n### But there's an even cheaper approach\n\nIf you have, say, **10,000 domains**, I wouldn't immediately pay for a traditional enrichment database.\n\nI'd build a **waterfall**:\n\n1. **Normalize the domain**\n2. Pull free/public company metadata:\n - website title/description\n - schema.org Organization data\n - OpenGraph metadata\n - LinkedIn/company URL where discoverable\n - favicon/logo\n - public About/Contact pages\n3. Use a cheap enrichment API **only when fields are missing**\n4. Cache every result permanently so you never pay twice\n5. Use a premium provider only for the few companies where you need difficult fields such as employee count, revenue, funding, or technology stack.\n\nThis can reduce your paid enrichment dramatically.\n\nFor comparison, People Data Labs currently lists Company Enrichment at **$0.10 per successful API request at its entry tier**, while Hunter's company lookup can consume just **0.2 credits when the requested company fields are returned**. [People Data Labs Help Center](https://support.peopledatalabs.com/hc/en-us/articles/25794271805211-Pricing-credits?utm_source=chatgpt.com)\n\n### My recommendation\n\nIf your requirement is specifically:\n\n> **\u201cGive me the cheapest way to take 1,000\u2013100,000 domains and turn them into company records, with no monthly subscription.\u201d**\n\nI'd start with **YepAPI or QuanticData at ~$30/1,000**, benchmark them against a small sample, and only move to PDL/Crustdata/etc. where their data is demonstrably better. [YepAPI](https://www.yepapi.com/email-api/company?utm_source=chatgpt.com)\n\nIf you tell me **how many domains you have and exactly which fields you need** (e.g. `employees, revenue, industry, country, LinkedIn, tech stack`), I can compare the **5\u201310 cheapest APIs and calculate the actual cost for your volume**.", "sources": [ "https://quanticdata.io/collectors/company-data-api/", "https://www.yepapi.com/email-api/company", "https://support.peopledatalabs.com/hc/en-us/articles/25794271805211-Pricing-credits", "https://you.com/resources/company-data-enrichment-api" ], "model": null, "provider_timestamp": null, "web_search": null, "search_queries": [], "served_by": null, "n_sources": 4, "grade": null, "brands": [ "Apollo", "People Data Labs", "Hunter", "ZoomInfo", "Crustdata" ] }, { "id": "P21", "route": "chatgpt-web", "latency_ms": 28558, "fetched_at": "2026-10-07T22:36:55Z", "answer": "For an **in-house growth agent**, I\u2019d optimize for APIs that let the agent *take actions and investigate opportunities programmatically*, rather than simply giving you a big SEO database.\n\n### My shortlist\n\n| API | Best for | Agent fit | My take |\n|---|---|---:|---|\n| **DataForSEO** | SERPs + keyword intelligence + backlinks | \u2b50\u2b50\u2b50\u2b50\u2b50 | **Best foundation** |\n| **Ahrefs API** | Backlinks + organic competitors + increasingly AEO | \u2b50\u2b50\u2b50\u2b50\u2b50 | **Best premium data layer** |\n| **Semrush API** | Broad SEO/competitive intelligence | \u2b50\u2b50\u2b50\u2b50 | **Best all-around enterprise suite** |\n| **SerpApi** | Raw Google/Bing/etc. SERPs | \u2b50\u2b50\u2b50\u2b50 | **Best SERP specialist** |\n| **Google Search Console API** | Your own site's actual performance | \u2b50\u2b50\u2b50\u2b50\u2b50 | **Essential, not optional** |\n| **Google Ads/Keyword Planner APIs** | Search volume / keyword demand | \u2b50\u2b50\u2b50\u2b50 | Great complement |\n| **Screaming Frog / custom crawler** | Technical SEO | \u2b50\u2b50\u2b50\u2b50 | Better as a crawler than an SEO data API |\n\n## 1. [DataForSEO](https://dataforseo.com/apis?utm_source=chatgpt.com) \u2014 my #1 for an agent\n\nIf you're actually **building an autonomous growth agent**, this is probably where I'd start.\n\nDataForSEO gives you separate APIs for:\n\n- Live SERPs\n- Keyword research\n- Search volume\n- Keyword difficulty\n- Search intent\n- Competitor discovery\n- Domain rankings\n- Backlinks\n- Google/Bing/YouTube/etc.\n- Historical keyword data\n\nIts Labs API, for example, exposes keyword, SERP, domain, competitor and ranking data, including historical search volume. [DataForSEO](https://docs.dataforseo.com/v3/dataforseo_labs/overview/?utm_source=chatgpt.com)\n\nIts SERP API is particularly useful because you can specify **keyword + search engine + language + location + device**, giving an agent a much more controllable search primitive. [DataForSEO](https://docs.dataforseo.com/v3/serp/overview/?utm_source=chatgpt.com)\n\nAnd the APIs are designed for programmatic consumption rather than forcing an agent through a human-oriented UI.\n\n**Why I like it for an agent:**\n\n```text\nUser/site\n \u2193\nGrowth Agent\n \u251c\u2500\u2500 GSC \u2192 \"What is actually happening?\"\n \u251c\u2500\u2500 DataForSEO SERP \u2192 \"What does Google show?\"\n \u251c\u2500\u2500 DataForSEO Labs \u2192 \"What opportunities exist?\"\n \u251c\u2500\u2500 DataForSEO Backlinks \u2192 \"Why do competitors win?\"\n \u2514\u2500\u2500 Crawler \u2192 \"What can we fix?\"\n \u2193\n Prioritization\n \u2193\n Content / SEO actions\n```\n\nIt also supports high request throughput\u2014DataForSEO documents up to 2,000 API calls/minute for its Live APIs, subject to concurrency limits. [DataForSEO](https://docs.dataforseo.com/v3/backlinks-overview/?utm_source=chatgpt.com)\n\n**Verdict: 9.5/10 for an in-house agent.**\n\n---\n\n## 2. [Ahrefs API](https://ahrefs.com/seo-api?utm_source=chatgpt.com) \u2014 best premium intelligence\n\nI'd choose Ahrefs if **backlinks, competitors, content opportunities and increasingly AI-search visibility** are central to your agent.\n\nIts current API exposes:\n\n- Site Explorer\n- Backlinks\n- Organic traffic\n- Keyword Explorer\n- SERP data\n- Rank Tracker\n- Site Audit\n- Batch analysis\n- Brand Radar / AI-search data\n\nThe interesting part for a 2026-era growth agent is **Brand Radar**: Ahrefs now exposes AI prompts and citations alongside traditional SEO data. [Ahrefs for Developers](https://docs.ahrefs.com/en/api/docs/introduction?utm_source=chatgpt.com)\n\nThat makes the data model increasingly:\n\n```text\nGoogle rankings\n +\nBacklinks\n +\nAI citations\n +\nAI prompts\n```\n\nrather than just traditional SEO.\n\nAhrefs also says its backlink index is refreshed every 15\u201330 minutes. [Ahrefs](https://ahrefs.com/seo-api?utm_source=chatgpt.com)\n\n**Verdict: 9/10.**\n\nIf budget isn't a major concern, I'd seriously consider **DataForSEO + Ahrefs** rather than trying to make one vendor do everything.\n\n---\n\n## 3. [Semrush API](https://developer.semrush.com/api/v4/?utm_source=chatgpt.com) \u2014 best broad enterprise option\n\nSemrush is attractive if your agent needs a very broad **marketing intelligence** layer.\n\nIts API ecosystem covers:\n\n- Organic keywords\n- Keyword research\n- Competitors\n- Backlinks\n- Traffic/market data\n- Position tracking\n- Site Audit projects\n- Local SEO\n- AI-search traffic data\n\nThe current API documentation also explicitly positions it for **AI workflows and internal applications**. [Semrush Developer](https://developer.semrush.com/api/v4/seo/overview/?utm_source=chatgpt.com)\n\nOne downside for an engineering-heavy agent is the commercial model: standard API access requires a Business subscription plus separately purchased API units. [Semrush Developer](https://developer.semrush.com/api/v3/get-started/api-access/?utm_source=chatgpt.com)\n\nSo I'd pick Semrush when you want:\n\n> \"Give my growth agent access to the entire Semrush intelligence ecosystem.\"\n\nI'd pick DataForSEO when you want:\n\n> \"Give my engineers composable SEO primitives at scale.\"\n\n**Verdict: 8.5/10.**\n\n---\n\n## 4. [SerpApi](https://serpapi.com/?utm_source=chatgpt.com) \u2014 excellent SERP primitive\n\nIf your agent's core loop is:\n\n> Search \u2192 inspect SERP \u2192 understand competitors \u2192 generate hypothesis \u2192 search again\n\nthen a dedicated SERP API is extremely useful.\n\nI'd particularly consider it if you want raw SERP observations rather than a traditional SEO suite.\n\nFor example:\n\n```text\n\"best accounting software\"\n \u2193\nSERP API\n \u2193\nPAA\nfeatured snippets\norganic results\nads\nlocal pack\nshopping\nknowledge panel\n \u2193\nAgent analyzes SERP\n \u2193\n\"What content format is Google rewarding?\"\n```\n\nI'd generally use **DataForSEO** instead if you also need extensive keyword/backlink/domain intelligence.\n\n---\n\n## 5. Google Search Console API \u2014 absolutely mandatory\n\nThis one is different.\n\nThird-party SEO APIs tell your agent what the **market** looks like.\n\nGSC tells it what is happening to **your actual site**.\n\nYour agent should be able to query things like:\n\n```text\nquery\npage\ncountry\ndevice\nclicks\nimpressions\nCTR\nposition\ndate\n```\n\nThen it can discover patterns such as:\n\n> \"This page gets 80k impressions/month for 37 queries but has a 1.8% CTR.\"\n\nThat's vastly more actionable than:\n\n> \"Keyword X has 12,000 monthly searches.\"\n\nFor an in-house agent, **first-party data should be the control plane**.\n\n---\n\n# What I'd actually build\n\nIf I were building your agent today, I'd use this stack:\n\n### Core\n\n**Google Search Console API**\n\n\u2192 actual performance\n\n**DataForSEO**\n\n\u2192 SERPs + keyword universe + competitors + search volume\n\n**Ahrefs**\n\n\u2192 backlinks + content/competitor intelligence + AEO\n\n**Google Analytics / product analytics**\n\n\u2192 conversion/business impact\n\n**Your own crawler**\n\n\u2192 technical/content inspection\n\nThen expose them to the agent as tools:\n\n```typescript\nseo.search_serp()\nseo.keyword_metrics()\nseo.keyword_opportunities()\nseo.competitor_keywords()\nseo.competitor_backlinks()\nseo.page_backlinks()\nseo.get_gsc_performance()\nseo.crawl_page()\nseo.crawl_site()\nseo.get_analytics()\n```\n\nThe important architectural decision is **not to expose 200 vendor endpoints to the LLM**.\n\nBuild a relatively small semantic layer.\n\nFor example:\n\n```text\nfind_growth_opportunities(site, market)\ndiagnose_ranking_drop(site, period)\nfind_quick_wins(site)\nanalyze_competitor(domain)\nfind_content_gaps(site, competitors)\nanalyze_serp(keyword)\nfind_pages_with_low_ctr(site)\nfind_pages_near_page_one(site)\n```\n\nThose become the agent's actual tools.\n\nThe vendor APIs become implementation details underneath.\n\n## My ranking for your use case\n\nIf your goal is specifically **\"an autonomous in-house growth agent\"**, I'd rank them:\n\n1. **DataForSEO** \u2014 best programmable SEO substrate\n2. **Ahrefs** \u2014 best premium SEO/AEO intelligence\n3. **Google Search Console** \u2014 indispensable first-party signal\n4. **Semrush** \u2014 excellent broad enterprise intelligence\n5. **SerpApi** \u2014 excellent if raw SERP retrieval is particularly important\n\nThe really interesting architecture is **DataForSEO + GSC + Ahrefs + your own crawler**, with the LLM sitting above them as the reasoning/orchestration layer.\n\nIf you tell me your **site type (SaaS/ecommerce/content/marketplace), approximate monthly organic traffic, and whether the agent should only recommend actions or actually execute them**, I can give you a concrete API stack\u2014including **which endpoints to buy, estimated monthly API cost, and the agent tool schema I'd use**.", "sources": [ "https://dataforseo.com/apis", "https://docs.dataforseo.com/v3/dataforseo_labs/overview/", "https://docs.dataforseo.com/v3/dataforseo_labs-google-overview/", "https://docs.dataforseo.com/v3/serp/overview/", "https://docs.dataforseo.com/v3/backlinks-overview/", "https://ahrefs.com/seo-api", "https://docs.ahrefs.com/en/api/docs/introduction", "https://developer.semrush.com/api/v4/", "https://developer.semrush.com/api/v3/get-started/api-access/", "https://developer.semrush.com/api/v4/seo/overview/", "https://developer.semrush.com/api/v3/introduction/semrush-api-overview/", "https://serpapi.com/" ], "model": null, "provider_timestamp": null, "web_search": null, "search_queries": [], "served_by": null, "n_sources": 12, "grade": null, "brands": [ "DataForSEO", "Ahrefs", "Semrush", "SerpApi", "Search Console" ] }, { "id": "P22", "route": "chatgpt-web", "latency_ms": 33138, "fetched_at": "2026-10-07T22:37:28Z", "answer": "Yes. You can get **most of the useful SEO data without paying for Ahrefs + Semrush + Moz simultaneously**. The trick is to use different free/cheap sources for different jobs.\n\n## A lean SEO toolkit\n\n### 1. Keyword volumes: Google Keyword Planner\n\n[Google Keyword Planner](https://support.google.com/google-ads/answer/7337243?hl=en______&utm_source=chatgpt.com) is probably the best free starting point.\n\nIt gives you:\n- Keyword ideas\n- Estimated monthly searches\n- Search trends\n- CPC\n- Geographic/language filtering\n\nGoogle says Keyword Planner can show estimates of how many searches keywords receive each month. [Google Help](https://support.google.com/google-ads/answer/7337243?hl=en______&utm_source=chatgpt.com)\n\n**Caveat:** Google requires you to set up a Google Ads account and enter billing information to access the basic keyword-discovery features. You don't necessarily need to run ads, though. [Google Help](https://support.google.com/google-ads/answer/7337243?hl=en______&utm_source=chatgpt.com)\n\n### 2. Keyword demand/trends: Google Trends\n\n[Google Trends](https://trends.google.com/home?utm_source=chatgpt.com) is excellent for figuring out whether a keyword is **growing, declining, seasonal, or basically dead**.\n\nIt's especially useful when Keyword Planner says two keywords have similar volume but you need to know which one has momentum. Trends lets you compare terms over time and against previous periods. [Google Help](https://support.google.com/trends/answer/17261722?hl=en&utm_source=chatgpt.com)\n\nFor example:\n\n> \"standing desk\" vs. \"adjustable desk\"\n\nYou might discover that their apparent search volumes don't tell the whole story.\n\n### 3. Your *real* keyword data: Google Search Console\n\n[Google Search Console](https://search.google.com/search-console/about?utm_source=chatgpt.com) is arguably more valuable than estimated keyword volume once your site has traffic.\n\nIt tells you:\n- Actual queries you're appearing for\n- Impressions\n- Clicks\n- CTR\n- Average position\n- Which pages rank for those queries\n\nGoogle specifically provides a Queries report showing the searches that brought users to your site. [Google Help](https://support.google.com/webmasters/answer/7576553?hl=en&utm_source=chatgpt.com)\n\nThis creates a great **\"keyword goldmine\"**:\n\n1. Export your GSC queries.\n2. Sort by impressions.\n3. Find keywords where you're ranking roughly positions 4\u201320.\n4. Improve those pages.\n5. Look for related queries you're getting impressions for but haven't explicitly targeted.\n\nYou don't need to know the estimated search volume to know that **Google is already sending impressions to your site for the query**.\n\n---\n\n## Backlinks without an Ahrefs subscription\n\n### 4. Ahrefs free tools\n\nThis is where things have improved considerably.\n\n[Ahrefs Free Backlink Checker](https://ahrefs.com/backlink-checker?target=webbsonline.com&utm_source=chatgpt.com) lets you check backlinks without subscribing to the full Ahrefs platform.\n\nMore interestingly, Ahrefs' current free account provides **Site Explorer for sites you verify ownership of**, including backlink/referring-domain reports and organic keyword data. The free account currently allows up to 1,000 backlinks/keywords visible at once. [Ahrefs Help Center](https://help.ahrefs.com/en/articles/13002606-what-can-i-use-for-free-in-ahrefs?utm_source=chatgpt.com)\n\nSo for **your own site**, you can get a surprisingly useful amount of data without paying for Ahrefs.\n\nFor competitor research, the free backlink checker gives you a sample rather than the complete database. [Ahrefs](https://ahrefs.com/backlink-checker?target=webbsonline.com&utm_source=chatgpt.com)\n\n### 5. Use Google's own link data for your site\n\nGoogle Search Console also has a **Links** report for your property. It's not a replacement for a commercial backlink index, but it's free and useful for identifying your important linking pages/domains.\n\nThe important distinction is:\n\n**You don't need a complete backlink database to do good SEO.**\n\nOften you mainly need to know:\n\n- Who links to competitors?\n- Which pages attract links?\n- What types of sites link to them?\n- Which pages on your site already attract links?\n- Where are obvious link opportunities?\n\nA sample of good backlinks can answer much of that.\n\n---\n\n# The setup I'd recommend\n\nIf I wanted to minimize subscriptions, I'd use:\n\n| Job | Tool | Cost |\n|---|---|---:|\n| Keyword ideas | Google Keyword Planner | Free |\n| Search trends | Google Trends | Free |\n| Your actual keywords | Google Search Console | Free |\n| Your backlinks | GSC + Ahrefs Free | Free |\n| Competitor backlink samples | Ahrefs Free Backlink Checker | Free |\n| SERP research | Google itself | Free |\n| Technical SEO | Screaming Frog free version | Free |\n| Analytics | Google Analytics | Free |\n\nThat gets you **a very capable SEO stack for $0/month**.\n\nThe main thing you give up versus Semrush/Ahrefs is **convenience and massive-scale competitor data**, not the ability to do SEO research.\n\n## One important trick\n\nDon't obsess over getting an exact number like:\n\n> \"This keyword has 8,100 searches/month.\"\n\nTreat volume as an **estimate**, not ground truth.\n\nI'd build a spreadsheet with:\n\n**Keyword | GKP volume | Trend | Current position | Impressions | CTR | Difficulty proxy | Target page**\n\nThen prioritize based on the combination rather than volume alone.\n\nFor example, a keyword with:\n\n- 1,500 estimated searches\n- 3,000 actual GSC impressions\n- position 8\n- 2.5% CTR\n\nmay be a **far better opportunity** than a supposedly 20,000-volume keyword where you're buried at position 70.\n\nIf you tell me **your niche/site and whether you're mainly trying to do keyword research, competitor research, or link building**, I can give you a specific **$0\u2013$20/month SEO workflow** that replaces most of what you'd normally use Ahrefs/Semrush for.", "sources": [ "https://support.google.com/google-ads/answer/7337243?hl=en______", "https://trends.google.com/home", "https://support.google.com/trends/answer/17261722?hl=en", "https://search.google.com/search-console/about", "https://support.google.com/webmasters/answer/7576553?hl=en", "https://help.ahrefs.com/en/articles/13002606-what-can-i-use-for-free-in-ahrefs", "https://ahrefs.com/backlink-checker?target=webbsonline.com" ], "model": null, "provider_timestamp": null, "web_search": null, "search_queries": [], "served_by": null, "n_sources": 7, "grade": null, "brands": [ "Ahrefs", "Semrush", "Moz", "Search Console" ] }, { "id": "P23", "route": "chatgpt-web", "latency_ms": 26299, "fetched_at": "2026-10-07T22:37:54Z", "answer": "If you only need **weekly keyword + Google SERP checks**, you can make this extremely cheap.\n\n### Cheapest practical setup\n\n**My pick: DataForSEO Standard SERP API.** It currently charges **$0.0006 per Google organic SERP (10 results)** on its standard queue\u2014about **$0.60 per 1,000 keyword checks**. The tradeoff is that results are queued rather than instant. [DataForSEO](https://dataforseo.com/pricing/google-serp/google-organic-serp-api?utm_source=chatgpt.com)\n\nFor example:\n\n| Keywords | Checks/week | Checks/month* | API cost/month |\n|---:|---:|---:|---:|\n| 100 | 100 | ~433 | **$0.26** |\n| 500 | 500 | ~2,167 | **$1.30** |\n| 1,000 | 1,000 | ~4,333 | **$2.60** |\n| 5,000 | 5,000 | ~21,667 | **$13.00** |\n| 10,000 | 10,000 | ~43,333 | **$26.00** |\n\n\\*Using 52 weeks / 12 months.\n\nYou don't need a full SEO rank-tracking SaaS for this. Your weekly job can simply:\n\n1. Store your keywords + target domain.\n2. Once a week, submit each keyword to the SERP API.\n3. Request only the **top 10 results** initially.\n4. Parse the results for your domain.\n5. Store `keyword \u2192 rank \u2192 URL \u2192 date`.\n6. Compare against last week's snapshot.\n7. Send yourself a report/alert.\n\nDataForSEO's API supports specifying the search location and language, and its default SERP depth is 10 results. [DataForSEO](https://docs.dataforseo.com/v3/serp-se-type-live-advanced/?utm_source=chatgpt.com)\n\n### One important wrinkle\n\nIf you want rankings beyond position 10, costs increase because DataForSEO bills additional 10-result chunks. So:\n\n- positions 1\u201310 \u2192 1 SERP unit\n- positions 1\u201320 \u2192 2 units\n- positions 1\u201350 \u2192 5 units\n- positions 1\u2013100 \u2192 10 units\n\n[DataForSEO](https://dataforseo.com/pricing/google-serp/google-organic-serp-api?utm_source=chatgpt.com)\n\nFor a weekly rank tracker, I'd actually consider **top 100** worthwhile, because otherwise a keyword disappearing from the top 10 looks like it vanished completely.\n\n### Even cheaper alternatives\n\nThere are newer pay-as-you-go SERP providers advertising around **$0.30\u2013$0.60/1,000 searches**, including serp.cheap, but I'd be more cautious about building a production workflow around a smaller provider. One current comparison puts serp.cheap at $0.30 cached / $0.60 fresh and DataForSEO at $0.60 standard. [serp.cheap](https://www.serp.cheap/blog/cheapest-serp-api/?utm_source=chatgpt.com)\n\nFor a **small weekly workload**, the absolute API difference is almost irrelevant. If you're checking 500 keywords weekly, you're talking about roughly **$1\u20132/month** in raw SERP costs.\n\n### What I would build\n\nFor example, for **500 keywords once per week**:\n\n```text\nScheduler (cron / GitHub Actions / Cloudflare Worker)\n \u2193\n DataForSEO API\n \u2193\n Parse top 100\n \u2193\n Find target domain\n \u2193\n PostgreSQL / SQLite\n \u2193\n \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2534\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510\n \u2193 \u2193\nRank history Weekly email\n```\n\nYou can keep the entire thing essentially serverless and probably spend **under $5/month total**, including the SERP API, depending on where you host it.\n\nIf you tell me roughly **how many keywords** you want to check (e.g. 100, 1,000, 10,000) and whether you need **top 10, top 100, or full SERP features**, I can calculate the cheapest architecture and exact monthly cost.", "sources": [ "https://dataforseo.com/pricing/google-serp/google-organic-serp-api", "https://docs.dataforseo.com/v3/serp-overview/", "https://docs.dataforseo.com/v3/serp-se-type-live-advanced/", "https://dataforseo.com/help-center/serp-api-pricing-depth-update-faq", "https://www.serp.cheap/blog/cheapest-serp-api/", "https://www.serp.cheap/blog/cheapest-google-search-api/" ], "model": null, "provider_timestamp": null, "web_search": null, "search_queries": [], "served_by": null, "n_sources": 6, "grade": null, "brands": [ "DataForSEO" ] }, { "id": "P25", "route": "chatgpt-web", "latency_ms": 32861, "fetched_at": "2026-10-07T22:38:27Z", "answer": "If your goal is **\u201cgive Claude Code access to my real SEO data and let it do keyword/content analysis from the terminal,\u201d** I\u2019d use an MCP-based setup rather than trying to make Claude scrape the Search Console UI.\n\n## My top choices\n\n| Tool | Best for | Claude Code fit | My take |\n|---|---|---:|---|\n| **Google Search Console API + MCP** | Your actual queries, clicks, impressions, CTR, rankings | \u2b50\u2b50\u2b50\u2b50\u2b50 | **Essential** |\n| **DataForSEO API/MCP** | Keyword volume, SERPs, competition, related keywords | \u2b50\u2b50\u2b50\u2b50\u2b50 | **Best API-first research layer** |\n| **Semrush API/MCP** | Broad competitive SEO intelligence | \u2b50\u2b50\u2b50\u2b50 | Best if you already pay for Semrush |\n| **Ahrefs API/MCP** | Backlinks + keyword/competitor research | \u2b50\u2b50\u2b50\u2b50 | Excellent, particularly for link analysis |\n| **Google Ads Keyword Planner API** | Google's keyword-volume estimates | \u2b50\u2b50\u2b50\u2b50 | Good inexpensive/official complement |\n| **Keywords Everywhere / similar APIs** | Lightweight keyword lookups | \u2b50\u2b50\u2b50 | Useful but less compelling for an agent |\n\n### 1. Google Search Console should be the foundation\n\nGoogle's Search Console API gives you programmatic access to Search Analytics, including **queries, pages, countries, devices, clicks, impressions, CTR and average position**. It also provides URL Inspection, sitemaps, and property management. [Google for Developers](https://developers.google.com/webmaster-tools/v1/api_reference_index?hl=en&utm_source=chatgpt.com)\n\nThat's particularly powerful with Claude Code because you can ask things like:\n\n> \"Find pages with 500+ impressions, position 4\u201315, and CTR below 3%. Suggest title/meta changes.\"\n\nor:\n\n> \"Find queries where we're getting impressions but have no dedicated page.\"\n\nor:\n\n> \"Compare the last 28 days with the previous 28 and identify content opportunities.\"\n\nThe API supports filtering/grouping by dimensions and up to 25,000 rows per request, with pagination. [Google for Developers](https://developers.google.com/webmaster-tools/v1/searchanalytics/query?utm_source=chatgpt.com)\n\n**Important:** Search Console isn't really a keyword-research database. It's **first-party performance data for your site**. That's why I'd pair it with a separate keyword/SERP data provider.\n\n[Google Search Console API documentation](https://developers.google.com/webmaster-tools/?utm_source=chatgpt.com)\n\n### 2. DataForSEO is probably my favorite keyword-research companion\n\nFor an agentic Claude Code workflow, I particularly like the **DataForSEO API** model because you can programmatically request things such as:\n\n- keyword search volume\n- CPC\n- competition\n- related keywords\n- SERPs\n- competitor rankings\n- keyword suggestions\n- historical data\n\nThat gives Claude the missing piece:\n\n**Search Console:** \"What are people actually finding us for?\"\n\n**DataForSEO:** \"What else do people search for, and what does Google currently rank?\"\n\nThat combination is much more useful than either alone.\n\n### 3. Semrush if you already use it\n\nIf you have a Semrush subscription, I'd seriously consider using its API rather than adding another SEO platform.\n\nIt gives Claude access to the sort of information you'd normally use Semrush for:\n\n- keyword research\n- competitor keywords\n- organic rankings\n- SERP information\n- backlinks\n- domain-level competitive analysis\n\nThe advantage isn't necessarily that it's *better* than every other API\u2014it is that Semrush gives you a very broad SEO dataset in one ecosystem.\n\n### 4. Ahrefs if backlinks are important\n\nI'd choose **Ahrefs** over Semrush if your workflow is particularly backlink/content-gap oriented.\n\nFor example, Claude could combine:\n\n```text\nSearch Console\n \u2193\nWhat we're already getting impressions for\n \u2193\nAhrefs\n \u2193\nWho ranks / what pages have backlinks\n \u2193\nClaude Code\n \u2193\nContent + internal-link + optimization recommendations\n```\n\nThat's a very strong workflow for an established site.\n\n---\n\n# What I'd actually build\n\nIf I were setting this up for Claude Code today, I'd use:\n\n```text\n \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510\n \u2502 Google Search \u2502\n \u2502 Console API \u2502\n \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u252c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518\n \u2502\n \u25bc\n\u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510 \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510\n\u2502 DataForSEO \u2502\u2500\u2500\u2500\u2500\u25b6\u2502 Claude Code \u2502\n\u2502 keywords/SERP \u2502 \u2502 \u2502\n\u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518 \u2502 SEO agent \u2502\n \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u252c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518\n \u2502\n \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2534\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510\n \u25bc \u25bc\n Site/codebase Content strategy\n```\n\nAnd expose them to Claude as MCP tools such as:\n\n```text\ngsc_search_analytics\ngsc_top_queries\ngsc_page_performance\ngsc_query_page_pairs\ngsc_inspect_url\n\nkeyword_volume\nkeyword_suggestions\nserp_results\ncompetitor_keywords\n```\n\nThen give Claude Code some higher-level commands/workflows:\n\n```text\n/analyze-seo-page URL\n/find-keyword-opportunities\n/find-content-gaps\n/find-ctr-opportunities\n/compare-search-performance\n/build-content-brief KEYWORD\n```\n\nThe **really interesting part** is combining Search Console's first-party data with external keyword data.\n\nFor example:\n\n```text\nGSC:\n\"best crm for startups\"\n1,842 impressions\nposition 8.7\nCTR 2.1%\n\n +\n\nKeyword database:\n18,000 monthly searches\n12 related queries\nSERP dominated by listicles\n\n \u2193\n\nClaude\n\n\"This page is an unusually strong optimization candidate.\nCreate a revised outline targeting the primary term plus these\n8 related entities/questions.\"\n```\n\nThat is considerably more useful than simply asking Claude to \"do keyword research.\"\n\n## One thing I'd avoid\n\nI wouldn't build the system around browser automation of Search Console.\n\nUse Google's API directly. Google requires OAuth 2.0 for Search Console API access, and the API is explicitly designed for programmatic access to the data. [Google for Developers](https://developers.google.com/webmaster-tools/v1/prereqs?utm_source=chatgpt.com)\n\nAlso, Search Console has quotas and data-return limitations, so I'd have the MCP layer cache/normalize results rather than having Claude repeatedly query six months of data. Google specifically recommends daily data retrieval patterns for comprehensive datasets. [Google for Developers](https://developers.google.com/webmaster-tools/limits?utm_source=chatgpt.com)\n\n### My ranking\n\n**If you're starting from scratch:**\n\n1. \ud83e\udd47 **Search Console API + custom MCP server**\n2. \ud83e\udd48 **DataForSEO**\n3. \ud83e\udd49 **Ahrefs or Semrush** \u2014 pick whichever you already subscribe to\n4. **Google Ads Keyword Planner API** as a supplementary source\n\nIf you tell me whether you want **free/cheap**, **best possible SEO data**, or **the simplest Claude Code setup**, I can give you a concrete MCP stack\u2014including the specific MCP servers/repos, installation commands, and a `CLAUDE.md` that turns Claude Code into an SEO research agent.", "sources": [ "https://developers.google.com/webmaster-tools/v1/searchanalytics/query", "https://developers.google.com/webmaster-tools/v1/api_reference_index?hl=en", "https://developers.google.com/webmaster-tools/", "https://developers.google.com/webmaster-tools/v1/prereqs", "https://developers.google.com/webmaster-tools/v1/how-tos/authorizing", "https://developers.google.com/webmaster-tools/limits", "https://developers.google.com/webmaster-tools/v1/how-tos/all-your-data" ], "model": null, "provider_timestamp": null, "web_search": null, "search_queries": [], "served_by": null, "n_sources": 7, "grade": null, "brands": [ "DataForSEO", "Ahrefs", "Semrush", "Search Console", "Keywords Everywhere" ] }, { "id": "P26", "route": "chatgpt-web", "latency_ms": 28455, "fetched_at": "2026-10-07T22:39:25Z", "answer": "If your goal is **programmatic competitor research across Meta + Google commercial ads**, the important thing to know is that the official APIs are not symmetric:\n\n- **Meta\u2019s official Ad Library API** is authoritative but is largely limited to political/social-issue ads, plus broader ad access for the UK/EU. It is **not a general commercial competitor-ad API**. [Facebook](https://br-fr.facebook.com/ads/library/api/?source=onboarding&utm_source=chatgpt.com)\n- **Google Ads API** is for accounts you manage, not competitor research. Google\u2019s **Ads Transparency Center has no documented public API for commercial ads**. [Google Help](https://support.google.com/google-ads/answer/15235?utm_source=chatgpt.com)\n\nSo for commercial competitive intelligence, I'd look at third-party APIs/scraping infrastructure.\n\n## My shortlist\n\n| API/provider | Meta | Google/YouTube | Historical data | Best for |\n|---|---|---|---|---|\n| **AdLibrary API** | \u2705 | \u2705 | \u2705 | Cross-platform ad-intelligence product |\n| **Apify** | \u2705 | \u2705 | Build yourself | Flexible/custom pipelines |\n| **Waldo** | \u2705 | \u2705 | \u2705 | Managed cross-library research |\n| **ScrapeCreators** | \u2705 | Some | Depends | Simple developer API |\n| **SerpApi-style solutions** | \u2014 | \u2705 | Limited | Google-focused research |\n\n### 1. AdLibrary API \u2014 best if you want a unified intelligence API\n\n[AdLibrary API](https://adlibrary.com/ad-library-api?utm_source=chatgpt.com)\n\nThis is probably the **closest fit to what you're describing**. It provides a single API across Facebook, Instagram, Google, YouTube and several other ad platforms, with filters for things like country, creative type, CTA, format, language and platform. [adlibrary.com](https://adlibrary.com/ad-library-api?utm_source=chatgpt.com)\n\nThe interesting part isn't merely \"get me ads.\" It's the ability to normalize competitive data across networks:\n\n```text\ncompetitor\n \u251c\u2500\u2500 Meta\n \u2502 \u251c\u2500\u2500 Facebook\n \u2502 \u2514\u2500\u2500 Instagram\n \u2514\u2500\u2500 Google\n \u251c\u2500\u2500 Search\n \u251c\u2500\u2500 Display\n \u2514\u2500\u2500 YouTube\n```\n\n**I'd choose this if:** you're building a competitor-monitoring SaaS, internal dashboard, agent, or recurring intelligence pipeline and don't want to maintain scrapers yourself.\n\n---\n\n### 2. Apify \u2014 best if you want control\n\n[Apify](https://apify.com/?utm_source=chatgpt.com)\n\nApify has separate actors for Meta Ad Library and Google's Ads Transparency Center. The Meta actors can return things like creative URLs, copy, CTA, landing URL, dates, platforms and ad-library IDs. [Apify](https://apify.com/viralanalyzer/facebook-ads-library?utm_source=chatgpt.com)\n\nFor Google, its Ads Transparency Center actors can retrieve advertiser/domain-based creatives, formats, dates and other available information. [Apify](https://apify.com/automation_craft/google-ads-transparency-scraper?utm_source=chatgpt.com)\n\n**I'd choose Apify if:**\n\n- you're comfortable owning the data pipeline;\n- you need custom extraction;\n- you want to store your own historical snapshots;\n- you want to combine ad data with website crawling, SERP data, social profiles, etc.\n\nThe downside is that you're effectively buying **scraping infrastructure rather than a pristine ad-intelligence dataset**. Google's interface in particular can change, so maintenance is part of the tradeoff.\n\n---\n\n### 3. Waldo \u2014 best managed cross-platform option\n\n[Waldo ad library API](https://www.waldo.fyi/use-cases/ad-library-api?utm_source=chatgpt.com)\n\nWaldo is interesting because it's explicitly designed around programmatic access to multiple public ad libraries. Its current offering covers Meta, Google, LinkedIn and TikTok, with scans stored over time. [Waldo](https://www.waldo.fyi/use-cases/ad-library-api?utm_source=chatgpt.com)\n\nThat makes it particularly attractive for **\"show me what Competitor X has been doing over the last 90 days\"** rather than simply **\"what ads are live right now?\"**\n\n---\n\n### 4. ScrapeCreators \u2014 good if you want a straightforward REST API\n\n[ScrapeCreators](https://scrapecreators.com/?utm_source=chatgpt.com)\n\nWorth considering if your priority is developer simplicity rather than building a whole scraping stack. Its Meta-focused approach is API-first and intended for pulling public Ad Library data into applications. [Scrape Creators](https://scrapecreators.com/blog/meta-ad-library-scraping?utm_source=chatgpt.com)\n\nI'd put it below AdLibrary/Apify for a serious multi-network competitive-intelligence system, but it can make sense for a relatively small internal tool.\n\n## What I'd actually build\n\nFor a **competitor research product**, I'd avoid thinking of this as simply an \"ad API.\"\n\nI'd build a normalized schema like:\n\n```text\nAdvertiser\n \u251c\u2500\u2500 platform\n \u251c\u2500\u2500 ad_id\n \u251c\u2500\u2500 first_seen\n \u251c\u2500\u2500 last_seen\n \u251c\u2500\u2500 status\n \u251c\u2500\u2500 creative\n \u2502 \u251c\u2500\u2500 image/video\n \u2502 \u251c\u2500\u2500 headline\n \u2502 \u251c\u2500\u2500 primary_text\n \u2502 \u2514\u2500\u2500 CTA\n \u251c\u2500\u2500 landing_page\n \u251c\u2500\u2500 destination_domain\n \u251c\u2500\u2500 country\n \u251c\u2500\u2500 placement\n \u2514\u2500\u2500 metadata\n```\n\nThen add your own derived intelligence:\n\n- **creative longevity** \u2014 ads running for 60+ days\n- **message clusters** \u2014 recurring hooks/offers\n- **creative fatigue** \u2014 replacement frequency\n- **landing-page changes**\n- **offer/pricing changes**\n- **Meta vs Google strategy**\n- **new competitor launches**\n- **creative format distribution**\n- **competitor \u2192 product \u2192 angle \u2192 CTA mapping**\n\nThat layer is where the real competitive advantage is. The raw libraries are increasingly commoditized.\n\n### My ranking\n\n**For most teams:**\n\n1. \ud83e\udd47 **AdLibrary API** \u2014 easiest cross-platform foundation\n2. \ud83e\udd48 **Waldo** \u2014 strong if persistent historical intelligence matters\n3. \ud83e\udd49 **Apify** \u2014 best flexibility/control\n4. **ScrapeCreators** \u2014 good lightweight developer option\n5. **Direct official APIs** \u2014 use where applicable, but don't expect them to solve commercial competitor research\n\nOne particularly important caveat: **don't confuse Google's Google Ads API with the Ads Transparency Center**. The former gives you data/actions for authorized advertiser accounts; it isn't a competitor-intelligence API. [Google Help](https://support.google.com/google-ads/answer/15235?utm_source=chatgpt.com)\n\nIf you're evaluating these for **building an internal competitor-research agent**, I can also compare **AdLibrary vs Waldo vs Apify vs SerpApi at the endpoint/field level, including estimated cost per 1,000 competitor ads and what historical data you actually get**.", "sources": [ "https://es-la.facebook.com/ads/library/api", "https://br-fr.facebook.com/ads/library/api/?source=onboarding", "https://www.itechguides.com/does-the-google-ads-transparency-center-have-an-api-whats-official-and-whats-not/", "https://support.google.com/google-ads/answer/15235", "https://adlibrary.com/ad-library-api", "https://apify.com/automation_craft/google-ads-transparency-scraper", "https://apify.com/brenton8907/google-ads-transparency-center/api", "https://apify.com/viralanalyzer/facebook-ads-library", "https://apify.com/", "https://www.waldo.fyi/use-cases/ad-library-api", "https://scrapecreators.com/", "https://scrapecreators.com/blog/meta-ad-library-scraping", "https://support.google.com/google-ads/answer/2375503?hl=en" ], "model": null, "provider_timestamp": null, "web_search": null, "search_queries": [], "served_by": null, "n_sources": 13, "grade": null, "brands": [ "SerpApi", "Apify", "ScrapeCreators", "Meta Ad Library", "Ads Transparency" ] }, { "id": "F1", "route": "chatgpt-web-nosearch", "latency_ms": 34280, "fetched_at": "2026-10-07T22:34:40Z", "answer": "The **Multimodal** filter in Google Search Console is a new way to see how your website performs when people find it through **image-based searches**, rather than traditional typed searches.\n\nGoogle introduced it in September 2026. [Google for Developers](https://developers.google.com/search/blog/2026/09/web-multimodal-in-sc?utm_source=chatgpt.com)\n\n### What counts as \u201cmultimodal\u201d?\n\nIt includes searches where the user provides an image as part of the search, such as:\n\n- **Google Lens** \u2014 someone points their phone camera at something and searches.\n- **Circle to Search** on Android \u2014 someone circles an object/image on their screen.\n- **Uploaded images** \u2014 someone uploads a photo to Google Search.\n- **Chrome \u201cSearch this image\u201d** \u2014 someone right-clicks an image and searches with it. [Google Help](https://support.google.com/webmasters/answer/7576553?hl=en&mbsy_source=1f1829&utm_source=chatgpt.com)\n\nSo, for example, imagine someone takes a photo of a particular type of chair. Google identifies it and shows your furniture article or product page in the results. **That impression/click can now be attributed to \u201cWeb: multimodal.\u201d**\n\n### Where you'll find it\n\nGo to:\n\n**Search Console \u2192 Performance \u2192 Search results \u2192 Search type \u2192 Web: Multimodal**\n\nGoogle now separates:\n\n- **Web: Text-based** \u2014 conventional searches where the user types a query.\n- **Web: Multimodal** \u2014 web searches where an image was used. [Google Help](https://support.google.com/webmasters/answer/7576553?hl=en&mbsy_source=1f1829&utm_source=chatgpt.com)\n\nYou can also see the multimodal category in the **Generative AI features** performance report. [Google Help](https://support.google.com/webmasters/answer/16984139?hl=en-GB&utm_source=chatgpt.com)\n\n### The important limitation: no query data\n\nThis is probably the most important thing to understand.\n\nWhen you select **Multimodal**, Search Console doesn't give you the actual search query. The **Queries** dimension isn't available because these searches primarily use an image rather than a text query. [Google Help](https://support.google.com/webmasters/answer/17011259?hl=en&trk=article-ssr-frontend-pulse_little-text-block&utm_source=chatgpt.com)\n\nInstead, you can analyze things such as:\n\n- Which **pages** receive multimodal impressions/clicks\n- Which **countries** they're coming from\n- Which **devices** are involved\n- How those metrics change over time\n\n### Why this matters for SEO\n\nThis is particularly useful if your site has lots of **visual content**, such as:\n\n- Ecommerce/product pages\n- Fashion\n- Furniture\n- Travel\n- Recipes/food\n- Architecture\n- Art\n- Automotive\n- How-to/tutorial content\n- Infographics and diagrams\n\nFor example, if one of your product pages gets **10,000 normal-search impressions but 2,000 multimodal impressions**, that's a signal that your visual content is being discovered through image-led search\u2014not just keywords.\n\nIt gives you a new way to measure **visual SEO** that previously wasn't separately visible in Search Console. Google says the data is being rolled out globally. [Google for Developers](https://developers.google.com/search/blog/2026/09/web-multimodal-in-sc?utm_source=chatgpt.com)\n\n**One subtle point:** don't interpret \u201cmultimodal\u201d as simply another version of Google Images. The **Image** search type represents traffic from results in Google Images, whereas **Web: multimodal** represents web results generated when an image was part of the search. [Google Help](https://support.google.com/webmasters/answer/7576553?hl=en&mbsy_source=1f1829&utm_source=chatgpt.com)\n\nIf you tell me what kind of website you have (ecommerce, blog, SaaS, local business, etc.), I can also explain **how I'd use the multimodal report to improve your SEO and what metrics to look for**.", "sources": [ "https://developers.google.com/search/blog/2026/09/web-multimodal-in-sc", "https://support.google.com/webmasters/answer/7576553?hl=en&mbsy_source=1f1829", "https://support.google.com/webmasters/answer/16984139?hl=en-GB", "https://support.google.com/webmasters/answer/17011259?hl=en&trk=article-ssr-frontend-pulse_little-text-block" ], "model": null, "provider_timestamp": null, "web_search": null, "search_queries": [], "served_by": null, "n_sources": 4, "grade": "correct", "brands": [ "Search Console" ] }, { "id": "F2", "route": "chatgpt-web-nosearch", "latency_ms": 18062, "fetched_at": "2026-10-07T22:34:58Z", "answer": "Yes. **Ahrefs increased its API limits in September 2026**, rather than reducing them.\n\nAhrefs\u2019 September 2026 product roundup says it **doubled the monthly API unit allowance on every plan**:\n\n| Plan | Before | After |\n|---|---:|---:|\n| Lite | 100,000 | **200,000** |\n| Standard | 400,000 | **800,000** |\n| Advanced | 1,000,000 | **2,000,000** |\n| Enterprise | 2,000,000 | **4,000,000** |\n\nAhrefs attributes the change partly to growing MCP usage. [Ahrefs](https://ahrefs.com/ja/blog/september-2026-roundup/?utm_source=chatgpt.com)\n\nThere is an interesting wrinkle: **Ahrefs' changelog dates the underlying API/MCP limit increase to May 27, 2026**, while its September roundup presents the doubled limits as a September update. [Ahrefs for Developers](https://docs.ahrefs.com/en/api/docs/changelog?utm_source=chatgpt.com) So if you're investigating a specific September billing/usage change, the date of the actual account-level change matters.\n\nAlso, the **60 requests/minute rate limit is separate from API-unit allowances** and is still documented as the default. [Ahrefs for Developers](https://docs.ahrefs.com/en/api/docs/introduction?utm_source=chatgpt.com)\n\nIf you're asking because **your API started returning 429s or consuming units differently in September**, tell me what changed (plan + endpoint/error), and I can dig into the exact September change.", "sources": [ "https://docs.ahrefs.com/en/api/docs/changelog", "https://ahrefs.com/ja/blog/september-2026-roundup/", "https://docs.ahrefs.com/en/api/docs/introduction" ], "model": null, "provider_timestamp": null, "web_search": null, "search_queries": [], "served_by": null, "n_sources": 3, "grade": "correct", "brands": [ "Ahrefs" ] }, { "id": "F3", "route": "chatgpt-web-nosearch", "latency_ms": 23836, "fetched_at": "2026-10-07T22:35:22Z", "answer": "OpenSEO is an **open-source, self-hostable alternative to SEO suites like Semrush and Ahrefs**. It focuses on practical SEO workflows rather than trying to reproduce every feature of a large enterprise suite. Its main capabilities include:\n\n- Keyword research and demand/difficulty estimates\n- Rank tracking and SERP features\n- Competitor/domain research\n- Backlink analysis\n- Technical site audits\n- AI-search/LLM visibility\n- An **MCP server**, so AI agents such as Claude Code, Codex, or OpenClaw can use the SEO data and workflows. [GitHub](https://github.com/afazhriel/openseo?utm_source=chatgpt.com)\n\n### What data source does it use?\n\nThe key point is that **OpenSEO isn't itself maintaining a giant proprietary SEO database**.\n\nIts primary underlying data provider is **DataForSEO**. OpenSEO calls DataForSEO's APIs for things such as keyword data, SERPs, rankings, backlinks, and domain information. The OpenSEO founder specifically says that discovering DataForSEO was what made the project possible. [OpenSEO](https://www.openseo.so/open-source-seo?utm_source=chatgpt.com)\n\nSo the architecture is roughly:\n\n**OpenSEO \u2192 DataForSEO \u2192 search/SEO data**\n\nFor self-hosting, you supply your own DataForSEO API credentials and pay DataForSEO according to usage; the OpenSEO software itself is free/open source. [OpenSEO](https://www.openseo.so/docs/self-hosting?utm_source=chatgpt.com)\n\nThere is also an important distinction with **Google Search Console**: OpenSEO can optionally connect to your GSC account, which gives it *your own first-party performance data* (queries, clicks, impressions, etc.). That's complementary to the broader third-party SEO data from DataForSEO. [OpenSEO](https://openseo.so/?utm_source=chatgpt.com)\n\nIf you're evaluating it against **Ahrefs/Semrush**, the interesting question is therefore not just \"is it open source?\" but **how DataForSEO's underlying keyword/backlink/SERP datasets compare in coverage and freshness to Ahrefs or Semrush's proprietary crawls**.", "sources": [ "https://openseo.so/", "https://github.com/afazhriel/openseo", "https://openseo.so/about", "https://www.openseo.so/open-source-seo", "https://www.openseo.so/docs/self-hosting" ], "model": null, "provider_timestamp": null, "web_search": null, "search_queries": [], "served_by": null, "n_sources": 5, "grade": "correct", "brands": [ "DataForSEO", "Ahrefs", "Semrush", "Search Console" ] }, { "id": "F4", "route": "chatgpt-web-nosearch", "latency_ms": 18048, "fetched_at": "2026-10-07T22:35:40Z", "answer": "Jev is [TypeSafe AI\u2019s](https://typesafe.ai/?utm_source=chatgpt.com) new AI model, but it\u2019s deliberately **not a chatbot/LLM**. TypeSafe calls it a **\u201cSystem One\u201d model**: instead of generating prose, Jev takes some context and returns a constrained, typed decision\u2014such as a choice, score, or yes/no probability\u2014with confidence information. [TypeSafe AI](https://typesafe.ai/blog/introducing-system-one-models-and-jev?utm_source=chatgpt.com)\n\nFor example, instead of asking an LLM, \u201cRead this support ticket and tell me what to do,\u201d you might give Jev the ticket and ask it to choose `billing`, `technical`, or `sales`. Your application can then branch directly on that result. The idea is essentially **\u201cAI as a very smart conditional/function\u201d** rather than AI as a text generator. [TypeSafe AI](https://typesafe.ai/blog/introducing-system-one-models-and-jev?utm_source=chatgpt.com)\n\nThe interesting part is the architecture/training objective. TypeSafe says Jev uses **Reinforcement Learning for Calibrated Decisions (RLCD)** and is optimized specifically for structured decisions rather than next-token text generation. Because it gives up arbitrary string generation, TypeSafe claims dramatically lower latency/cost on these bounded decision tasks. Those performance numbers should be treated as vendor claims unless independently reproduced. [TypeSafe AI](https://typesafe.ai/blog/introducing-system-one-models-and-jev?utm_source=chatgpt.com)\n\nSo conceptually:\n\n**LLM:** context \u2192 reasoning/text \u2192 `\"I think you should route this to billing...\"`\n\n**Jev:** context + predefined question \u2192 `billing: 0.96` \u2192 your code acts.\n\nIt was introduced in September 2026 and is named after economist **William Stanley Jevons**, referencing Jevons' paradox: making intelligence much cheaper could cause its total use to explode. [TypeSafe AI](https://typesafe.ai/blog/introducing-system-one-models-and-jev?utm_source=chatgpt.com)\n\nThe genuinely interesting question is whether Jev is **actually a fundamentally different model from an LLM**, or basically a highly optimized classifier/structured-output model with clever training. I can break down what appears to be technically novel about Jev versus ordinary LLM structured outputs if that's what you're curious about.", "sources": [ "https://typesafe.ai/", "https://typesafe.ai/blog/introducing-system-one-models-and-jev" ], "model": null, "provider_timestamp": null, "web_search": null, "search_queries": [], "served_by": null, "n_sources": 2, "grade": "correct", "brands": [] }, { "id": "P02", "route": "chatgpt-web-nosearch", "latency_ms": 22612, "fetched_at": "2026-10-07T22:38:12Z", "answer": "If you mean **\u201cI have a list of company domains and want firmographic data for each, but I don\u2019t want to pay a recurring SaaS subscription,\u201d** the cheapest options are usually **pay-as-you-go APIs** or free tiers.\n\n### Cheapest options I\u2019d look at\n\n| Option | Pricing model | Approx. cost | Best for |\n|---|---|---:|---|\n| **YepAPI** | Pay per call | **$0.03/domain** | Simple, occasional enrichment |\n| **People Data Labs** | Pay/credit based | From **$0.01/record** | Larger datasets / richer company data |\n| **Diffbot** | Free tier + usage | **10,000 credits/month free** | Lots of enrichment, if their fields fit |\n| **Hunter** | Free + prepaid/API options | Free tier available | Basic company/email information |\n| **Apollo** | Free credits, then subscription | Free tier | Occasional enrichment, if credits suffice |\n\nYepAPI explicitly advertises **$0.03 per company lookup with no monthly minimum or subscription**, and gives $5 in initial credit. [YepAPI](https://www.yepapi.com/email-api/company?utm_source=chatgpt.com)\n\nPeople Data Labs is potentially much cheaper at scale: its company enrichment API consumes one credit per successful enrichment, with published pricing starting around **$0.10/record** on its lower self-serve tier, although some current comparisons report pay-as-you-go options around $0.01/record depending on product/volume. [People Data Labs Help Center](https://support.peopledatalabs.com/hc/en-us/articles/25794271805211-Pricing-credits?utm_source=chatgpt.com)\n\nDiffbot is particularly interesting if you need **thousands of domains**: current comparisons list **10,000 free credits/month**, though the exact number of credits consumed depends on the API operation. [Datamagnet](https://www.datamagnet.co/post/company-enrichment-apis-compared/?utm_source=chatgpt.com)\n\n### My recommendation\n\nIf you're doing something like:\n\n> **10,000 domains \u2192 company name, industry, employee count, location, revenue, LinkedIn, etc.**\n\nI'd **not buy Apollo/ZoomInfo/Clearbit/Clay just for this**.\n\nI'd build a tiny enrichment pipeline:\n\n**CSV of domains \u2192 API \u2192 JSON/CSV \u2192 deduplicate/cache results**\n\nFor occasional batches, **YepAPI** is probably the simplest because there's genuinely no monthly commitment. At 10,000 domains, its advertised rate would be about **$300**. [YepAPI](https://www.yepapi.com/email-api/company?utm_source=chatgpt.com)\n\nFor repeated/high-volume enrichment, I'd benchmark **People Data Labs vs. Diffbot vs. a pay-as-you-go provider** based on the *fields you actually need*, because the cheapest API isn't necessarily the cheapest **per usable/enriched company**.\n\nIf you tell me **how many domains you need to enrich (e.g. 1K, 10K, 100K) and which fields you need**, I can compare the cheapest options and give you an estimated total cost.", "sources": [ "https://www.yepapi.com/email-api/company", "https://www.enrich.so/alternatives/clearbit", "https://support.peopledatalabs.com/hc/en-us/articles/25794271805211-Pricing-credits", "https://www.datamagnet.co/post/company-enrichment-apis-compared/" ], "model": null, "provider_timestamp": null, "web_search": null, "search_queries": [], "served_by": null, "n_sources": 4, "grade": null, "brands": [ "Apollo", "Clearbit", "People Data Labs", "Hunter", "ZoomInfo", "Clay" ] }, { "id": "P21", "route": "chatgpt-web-nosearch", "latency_ms": 36247, "fetched_at": "2026-10-07T22:36:16Z", "answer": "For an **in-house growth agent**, I\u2019d optimize for APIs that let an agent *decide and act*, rather than APIs that merely reproduce an SEO dashboard.\n\nMy shortlist:\n\n| API | Best for | Agent value | My take |\n|---|---|---:|---|\n| **[Ahrefs API](https://ahrefs.com/seo-api?utm_source=chatgpt.com)** | Backlinks, keywords, competitors, SERPs, AEO | \u2b50\u2b50\u2b50\u2b50\u2b50 | **Best all-around SEO intelligence layer** |\n| **[DataForSEO](https://dataforseo.com/?utm_source=chatgpt.com)** | Raw SERPs, keyword research, backlinks, scalable querying | \u2b50\u2b50\u2b50\u2b50\u2b50 | **Best API-native / programmable choice** |\n| **[Semrush API](https://developer.semrush.com/api/?utm_source=chatgpt.com)** | Competitive research, keywords, backlinks, traffic/market data | \u2b50\u2b50\u2b50\u2b50\u00bd | **Best broad marketing intelligence** |\n| **[Google Search Console API](https://developers.google.com/webmaster-tools/v1/api_reference_index?utm_source=chatgpt.com)** | Your actual impressions, clicks, queries, pages | \u2b50\u2b50\u2b50\u2b50\u2b50 | **Essential first-party data** |\n| **[Google Ads API](https://developers.google.com/google-ads/api/docs/keyword-planning/overview?utm_source=chatgpt.com)** | Search demand, keyword ideas, forecasts | \u2b50\u2b50\u2b50\u2b50 | Excellent demand signal |\n| **[Majestic API](https://developer.majestic.com/?utm_source=chatgpt.com)** | Link graph/backlink intelligence | \u2b50\u2b50\u2b50 | Great specialist supplement |\n\n### 1. Ahrefs \u2014 my default choice\n\nAhrefs has become particularly interesting for an **AI growth agent** because its API now spans Site Explorer, Keywords Explorer, SERPs, Rank Tracker, Site Audit, and **Brand Radar/AEO data**. Its current API also exposes real AI prompts and citations, not just traditional Google SEO metrics. [Ahrefs](https://ahrefs.com/seo-api?utm_source=chatgpt.com)\n\nThat makes it useful for agent loops such as:\n\n> \"Find pages where we're losing traffic \u2192 identify the competing pages \u2192 determine why they win \u2192 find related queries \u2192 recommend/update content \u2192 monitor whether the change worked.\"\n\nThe backlink index is also refreshed every 15\u201330 minutes according to Ahrefs. [Ahrefs](https://ahrefs.com/seo-api?utm_source=chatgpt.com)\n\n**I'd choose Ahrefs if:** you want one relatively comprehensive intelligence source and care about **SEO + AEO/AI visibility**.\n\n---\n\n### 2. DataForSEO \u2014 arguably the best *engineering* choice\n\nThis is the one I'd seriously consider if you're actually building an autonomous agent rather than an internal dashboard.\n\nDataForSEO gives you relatively granular APIs for:\n\n- Google/Bing/etc. SERPs\n- keyword research\n- historical SERPs\n- search intent\n- backlinks\n- domains\n- on-page analysis\n- business/local data\n- AI/LLM mentions\n\nIts SERP API supports Google, Bing, YouTube, Yahoo, Baidu, Naver and others, with explicit location, language, device and search-engine parameters. [DataForSEO Documentation](https://docs.dataforseo.com/v3/serp-overview/?utm_source=chatgpt.com)\n\nAnd its Google Organic SERP pricing is currently around **$0.002 per 10-result SERP** at the base rate, before additional parameters/depth. [DataForSEO](https://dataforseo.com/pricing/serp/google-organic-serp-api?utm_source=chatgpt.com)\n\nThat's very attractive for an agent that might make **thousands or millions of small queries**.\n\nTheir Labs API also has keyword/competitor research and a 4.8B+ keyword database. [DataForSEO Documentation](https://docs.dataforseo.com/v3/dataforseo_labs-google-overview/?utm_source=chatgpt.com)\n\n**I'd choose DataForSEO if:** your engineering team wants maximum control, high query volume, predictable pay-as-you-go economics, and raw search data.\n\n---\n\n### 3. Semrush \u2014 best if growth extends beyond SEO\n\nSemrush has unusually broad coverage. Its API ecosystem currently includes:\n\n- SEO / keyword data\n- backlinks\n- domain/competitor research\n- Projects/Site Audit\n- traffic & market intelligence\n- local SEO\n- AI-search traffic data\n\nIts current API documentation explicitly describes AI search traffic from ChatGPT, Gemini and Copilot, alongside traditional SEO data. [Semrush Developer](https://developer.semrush.com/api/v3/introduction/semrush-api-overview/?utm_source=chatgpt.com)\n\nThe downside is that API access is more tightly coupled to Semrush subscriptions/API units. [Semrush Developer](https://developer.semrush.com/api/v3/get-started/api-access/?utm_source=chatgpt.com)\n\n**I'd choose Semrush if:** your \"growth agent\" will eventually reason across **SEO + paid search + competitive intelligence + traffic/market data**.\n\n---\n\n## The important one: Google Search Console\n\nI wouldn't build the agent entirely on third-party SEO data.\n\nGive it **GSC as the ground truth for your own site**.\n\nThe agent should continuously ingest:\n\n```text\nquery\npage\ncountry\ndevice\ndate\nclicks\nimpressions\nCTR\naverage_position\n```\n\nThen combine that with external intelligence.\n\nFor example:\n\n```text\nGSC:\n\"best crm for startups\"\n\u2192 42,000 impressions\n\u2192 1.8% CTR\n\u2192 position 5.7\n\nAhrefs/DataForSEO:\n\u2192 competitors ranking 1\u20134\n\u2192 their referring domains\n\u2192 SERP features\n\u2192 related queries\n\u2192 content characteristics\n\nAgent:\n\u2192 \"This is a high-impression CTR opportunity.\"\n\u2192 \"Competitor #2 has 3x referring domains.\"\n\u2192 \"Our title doesn't match the dominant intent.\"\n\u2192 \"Rewrite title/meta + add comparison section.\"\n```\n\nThat's much more powerful than simply asking an LLM to \"do SEO.\"\n\n---\n\n## Google Ads API is underrated\n\nI'd also connect the **Google Ads Keyword Planning API**.\n\nIt can generate keyword ideas from:\n\n- keywords\n- URLs\n- entire sites\n\nand return historical search metrics. It can also generate forecasts. [Google for Developers](https://developers.google.com/google-ads/api/docs/keyword-planning/overview?utm_source=chatgpt.com)\n\nThat gives your agent an additional signal for **actual search demand**, rather than relying exclusively on Ahrefs/Semrush's modeled volume.\n\n---\n\n# What I'd actually build\n\nIf I were designing an in-house growth agent today, I'd use:\n\n```text\n \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510\n \u2502 Growth Agent \u2502\n \u2502 \u2502\n \u2502 Research \u2502\n \u2502 Diagnose \u2502\n \u2502 Prioritize \u2502\n \u2502 Recommend \u2502\n \u2502 Execute \u2502\n \u2502 Measure \u2502\n \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u252c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518\n \u2502\n \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u253c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510\n \u2502 \u2502 \u2502\n \u25bc \u25bc \u25bc\n Google GSC Ahrefs / DataForSEO\n first-party Semrush raw search\n \u2502 \u2502 \u2502\n \u2502 \u2502 \u2502\n \u25bc \u25bc \u25bc\n Actual traffic Competitors Live SERPs\n Actual queries Backlinks SERP features\n Actual CTR Keywords Locations\n Actual rankings AEO Devices\n \u2502\n \u25bc\n Google Ads API\n Search demand\n Forecasts\n```\n\n### My preferred stack\n\n**Lean version:**\n\n> **GSC + DataForSEO + Google Ads API**\n\nThis is probably the best **cost/control** combination.\n\n**Premium version:**\n\n> **GSC + Ahrefs + Google Ads API**\n\nThis is what I'd pick if SEO/AEO is strategically important and API spend isn't the primary constraint.\n\n**Broad growth intelligence version:**\n\n> **GSC + Semrush + DataForSEO + Google Ads**\n\nThis gives the agent both high-level competitive intelligence and granular SERP access, but there's more overlap.\n\n---\n\n## One architectural recommendation\n\nDon't let the LLM call these APIs arbitrarily.\n\nPut a **normalized SEO data layer** in front of them:\n\n```text\n \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510\n \u2502 Growth LLM \u2502\n \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u252c\u2500\u2500\u2500\u2500\u2500\u2500\u2518\n \u2502\n \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u25bc\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510\n \u2502 SEO Data Tools \u2502\n \u2502 \u2502\n \u2502 keyword() \u2502\n \u2502 serp() \u2502\n \u2502 competitors() \u2502\n \u2502 backlinks() \u2502\n \u2502 rankings() \u2502\n \u2502 opportunities()\u2502\n \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u252c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518\n \u2502\n \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u253c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510\n \u25bc \u25bc \u25bc\n GSC Ahrefs DataForSEO\n```\n\nThat gives you **vendor independence** and lets you swap Ahrefs \u2192 Semrush or DataForSEO without retraining/rearchitecting the agent.\n\nI'd also store the raw observations in your warehouse. SEO agents become dramatically more useful when they can reason about **change over time**, rather than querying today's SERP every time.\n\n### Bottom line\n\nIf you ask me to pick **one**:\n\n**\u2192 Ahrefs for the best overall SEO/AEO intelligence.**\n\nIf you ask me to pick **one for an engineering-heavy autonomous agent**:\n\n**\u2192 DataForSEO.**\n\nIf you're building a serious in-house growth system:\n\n**\u2192 GSC + DataForSEO + Ahrefs** is the combination I'd start with.\n\nThe particularly interesting 2026 shift is that **AEO/AI-search visibility is becoming a first-class data source**. Ahrefs' Brand Radar and Semrush's AI-search data make them considerably more interesting for an agent than they were when these APIs were primarily keyword/backlink databases. [Ahrefs](https://ahrefs.com/seo-api?utm_source=chatgpt.com)", "sources": [ "https://ahrefs.com/seo-api", "https://dataforseo.com/", "https://developer.semrush.com/api/", "https://developers.google.com/webmaster-tools/v1/api_reference_index", "https://developers.google.com/google-ads/api/docs/keyword-planning/overview", "https://docs.ahrefs.com/en/api/docs/introduction", "https://dataforseo.com/pricing/serp/google-organic-serp-api", "https://docs.dataforseo.com/v3/serp-overview/", "https://docs.dataforseo.com/v3/dataforseo_labs-google-overview/", "https://developer.semrush.com/api/v3/introduction/semrush-api-overview/", "https://developer.semrush.com/api/v3/get-started/api-access/", "https://developers.google.com/google-ads/api/docs/keyword-planning/generate-keyword-ideas" ], "model": null, "provider_timestamp": null, "web_search": null, "search_queries": [], "served_by": null, "n_sources": 12, "grade": null, "brands": [ "DataForSEO", "Ahrefs", "Semrush", "Search Console", "Majestic" ] }, { "id": "P22", "route": "chatgpt-web-nosearch", "latency_ms": 28014, "fetched_at": "2026-10-07T22:36:44Z", "answer": "Yes. You can get **most of the useful SEO data you need without maintaining several $100+/month subscriptions**. The trick is to use different free/cheap sources for different jobs rather than expecting one tool to do everything.\n\n## A lean SEO stack\n\n| Need | Low-cost option | What you get |\n|---|---|---|\n| **Keyword volumes** | [Semrush Keyword Search Volume Checker](https://www.semrush.com/free-tools/keyword-search-volume-checker/?utm_source=chatgpt.com) | Free individual volume checks, difficulty, intent, CPC |\n| **Keyword ideas + volumes** | [Google Keyword Planner](https://support.google.com/google-ads/answer/7337243?hl=en______&utm_source=chatgpt.com) | Keyword ideas and estimated monthly searches |\n| **Your actual search data** | Google Search Console | Impressions, clicks, CTR, queries and positions for your own site |\n| **Your backlinks** | [Ahrefs Free](https://ahrefs.com/free?utm_source=chatgpt.com) | Up to 1,000 backlinks/keywords for verified sites |\n| **Competitor backlinks** | [Ahrefs Free Backlink Checker](https://ahrefs.com/backlink-checker?target=webbsonline.com&utm_source=chatgpt.com) | Limited free backlink sample for any domain |\n| **Competitor keyword research** | Free tools + occasional paid month | Avoid keeping a subscription running continuously |\n\nGoogle Keyword Planner is particularly useful because it provides monthly-search estimates and keyword ideas, although Google currently requires billing information to access its basic Keyword Planner functionality. [Google Help](https://support.google.com/google-ads/answer/3022575?hl=en&utm_source=chatgpt.com)\n\n### The biggest money-saving trick\n\n**Don't pay for Ahrefs/Semrush/Moz continuously.**\n\nInstead, I'd use a **\"subscription burst\" model**:\n\n1. **Most months:** use Google Search Console + free keyword/backlink tools.\n2. **Every 2\u20133 months:** subscribe to *one* premium tool for a month.\n3. Export your keyword/backlink/competitor research.\n4. Cancel it.\n5. Work from your spreadsheet until you need another data refresh.\n\nFor your own site, Ahrefs Free is surprisingly capable now: verified sites can see up to **1,000 backlinks and 1,000 organic keywords**, with unlimited verified projects. [Ahrefs Help Center](https://help.ahrefs.com/en/articles/13002606-what-can-i-use-for-free-in-ahrefs?utm_source=chatgpt.com)\n\n### For backlinks specifically\n\nI'd separate **\"finding backlinks\"** from **\"finding link opportunities.\"**\n\nFor your own site:\n\n- Ahrefs Free \u2192 backlink profile.\n- Google Search Console \u2192 Google's view of linking sites.\n- Manually record referring domains, target pages and anchor text.\n\nFor competitors:\n\n- Use Ahrefs' free Backlink Checker to identify their most interesting referring domains/pages.\n- Search Google for things like `\"competitor name\"`, resource pages, industry directories, guest contributions, statistics pages, etc.\n- You don't actually need 50,000 backlink rows. **The first 50\u2013200 genuinely relevant referring domains are usually much more actionable.**\n\nAhrefs' current free offering explicitly supports backlink analysis for verified sites, while its standalone Backlink Checker provides a limited sample for arbitrary domains. [Ahrefs](https://ahrefs.com/webmaster-tools/?utm_source=chatgpt.com)\n\n## What I'd personally use\n\nIf you're doing SEO for **one or a handful of sites**, I'd start with:\n\n**$0/month**\n- Google Search Console\n- Google Keyword Planner\n- Semrush's free volume checker\n- Ahrefs Free\n- Google itself/SERPs\n- A spreadsheet\n\nThen, when you need serious competitor research, **buy one month of Ahrefs or Semrush**, do a big export, and cancel.\n\nThat can turn a **$200\u2013$500/month SEO-tool stack into roughly $0\u2013$50/month averaged over the year**, depending on how frequently you need fresh competitive data.\n\nIf you tell me **what kind of SEO you're doing (affiliate, local business, SaaS, ecommerce, agency, etc.) and roughly how many sites**, I can lay out a specific **$0 / $20 / $50-per-month stack**, including exactly where I'd get keyword volume, KD, SERPs, backlinks, DR/DA and competitor data.", "sources": [ "https://support.google.com/google-ads/answer/7337243?hl=en______", "https://www.semrush.com/free-tools/keyword-search-volume-checker/", "https://ahrefs.com/free", "https://ahrefs.com/backlink-checker?target=webbsonline.com", "https://support.google.com/google-ads/answer/3022575?hl=en", "https://help.ahrefs.com/en/articles/13002606-what-can-i-use-for-free-in-ahrefs", "https://ahrefs.com/webmaster-tools?ssrid=ssr", "https://ahrefs.com/webmaster-tools/" ], "model": null, "provider_timestamp": null, "web_search": null, "search_queries": [], "served_by": null, "n_sources": 8, "grade": null, "brands": [ "Ahrefs", "Semrush", "Moz", "Search Console" ] }, { "id": "P23", "route": "chatgpt-web-nosearch", "latency_ms": 29366, "fetched_at": "2026-10-07T22:37:13Z", "answer": "If you mean **weekly Google keyword/rank checks via an API**, the cheapest practical option I\u2019d use is **DataForSEO\u2019s SERP API**, assuming you don\u2019t need results instantly.\n\n### Cheapest options\n\n| API | Approx. cost / 1,000 Google SERPs | Best for |\n|---|---:|---|\n| **DataForSEO Standard** | **$0.60** | Cheapest overall |\n| **Serper** | $1.00 down to $0.30 at high volume | Simple, fast API |\n| **DataForSEO Live** | $2.00 | Real-time checks |\n| **SerpApi** | ~$25+/mo entry plan | Easier developer experience |\n\nDataForSEO currently charges **$0.0006 per Google Organic SERP** on its standard queue, with roughly 5-minute average turnaround. Its Live API is $0.002/SERP. [DataForSEO](https://dataforseo.com/pricing/google-serp/google-organic-serp-api?utm_source=chatgpt.com)\n\n[DataForSEO SERP API pricing](https://dataforseo.com/pricing/google-serp/google-organic-serp-api?utm_source=chatgpt.com)\n\nSerper is also attractive if you want a very simple synchronous API: its published pricing starts at **$1 per 1,000 queries**, falling to $0.30/1,000 at its largest credit tier. [Serper](https://serper.dev/?trk=public_post-text&utm_source=chatgpt.com)\n\n[Serper API](https://serper.dev/?utm_source=chatgpt.com)\n\n### What that means for weekly tracking\n\nSuppose you have **1,000 keywords**, checking each keyword once per week:\n\n- 1,000 keywords \u00d7 52 weeks = **52,000 SERPs/year**\n- DataForSEO: about **$31/year**\n- Serper at $1/1k: about **$52/year**\n- Even 10,000 keywords weekly would only be about **$312/year** with DataForSEO's standard rate.\n\nThat's just the raw SERP lookup cost; your scheduler/database/server can be essentially negligible.\n\n### The architecture I'd recommend\n\nFor a cheap rank tracker:\n\n```text\nWeekly cron\n \u2193\nKeyword list\n \u2193\nDataForSEO Standard API\n \u2193\nGoogle SERP JSON\n \u2193\nExtract your domain's position\n \u2193\nStore: keyword + date + position + URL\n \u2193\nCompare with previous week\n```\n\nYou **don't need a dedicated rank-tracking service** if all you want is weekly positions. Query Google for each keyword, inspect the organic results, and record the first result matching your domain.\n\nOne important cost optimization: **don't request 100 results if you only care whether you're in the top 10/20/30**. DataForSEO bills based on SERP depth, so keeping the requested depth low materially reduces cost. [DataForSEO](https://dataforseo.com/help-center/serp-api-cost-explained?utm_source=chatgpt.com)\n\nIf you tell me roughly **how many keywords you have (e.g. 100, 1,000, 10,000)** and whether you need **country/city-specific rankings**, I can calculate the cheapest setup and compare DataForSEO vs. Serper vs. building a scraper yourself. [DataForSEO](https://dataforseo.com/pricing/google-serp/google-organic-serp-api?utm_source=chatgpt.com)", "sources": [ "https://dataforseo.com/pricing/google-serp/google-organic-serp-api", "https://dataforseo.com/help-center/serp-api-cost-explained", "https://serper.dev/", "https://serper.dev/?trk=public_post-text" ], "model": null, "provider_timestamp": null, "web_search": null, "search_queries": [], "served_by": null, "n_sources": 4, "grade": null, "brands": [ "DataForSEO", "SerpApi", "Serper" ] }, { "id": "P25", "route": "chatgpt-web-nosearch", "latency_ms": 35862, "fetched_at": "2026-10-07T22:37:49Z", "answer": "If your goal is **\u201cgive Claude Code live access to my Search Console + real keyword/competitor data, so it can actually do SEO work\u201d**, I\u2019d use an MCP-based stack rather than CSV exports.\n\n## My picks\n\n| Tool | Best for | Claude Code | My take |\n|---|---|---|---|\n| **Google Search Console MCP** | Your actual clicks, impressions, queries, pages, CTR, indexing | \u2705 | **Essential** |\n| **Ahrefs MCP** | Keyword research, competitors, backlinks, traffic potential | \u2705 | **Best all-around paid option** |\n| **Semrush MCP** | Keyword research + competitive/market intelligence | \u2705 | **Best alternative to Ahrefs** |\n| **DataForSEO MCP/API** | SERPs, keyword data, programmatic SEO research | \u2705 | **Best for automation/control** |\n| **Google Ads Keyword Planner/API** | Google's own keyword-volume ecosystem | Via API/custom MCP | Useful, but less convenient |\n\n### 1. Google Search Console MCP \u2014 absolutely get this\n\nGSC answers the question that third-party keyword tools can't:\n\n> **\u201cWhat is Google actually showing my site for?\u201d**\n\nYou can expose query, page, country, device, clicks, impressions, CTR and position data directly to Claude. There are several community MCP implementations; one particularly interesting current option is **Luke Renton's `google-search-console-mcp`**, which is explicitly designed for Claude Code and includes URL inspection and sitemap functionality in addition to Search Analytics. [GitHub](https://github.com/LukeRenton/google-search-console-mcp?utm_source=chatgpt.com)\n\nFor example, Claude can then answer:\n\n> \u201cFind pages getting 1,000+ impressions/month where we're ranking positions 4\u201315, identify the queries responsible, and recommend which pages I should update.\u201d\n\nThat's considerably more useful than asking Claude to analyze a GSC CSV.\n\n### 2. Ahrefs MCP \u2014 my first choice for keyword research\n\n[Ahrefs MCP](https://ahrefs.com/mcp/?utm_source=chatgpt.com) is probably the cleanest complement to GSC.\n\nIt gives Claude access to things such as:\n\n- keyword volume\n- keyword difficulty\n- traffic potential\n- keyword trends\n- competitor keywords\n- organic traffic estimates\n- backlinks\n- competitor gaps\n\nAhrefs specifically supports Claude and makes its MCP available on paid plans starting with Lite. [Ahrefs](https://ahrefs.com/mcp/?utm_source=chatgpt.com)\n\nThe combination is powerful:\n\n**GSC = what you're already winning/losing**\n\n**Ahrefs = what the market is doing + what you could target**\n\nThat lets Claude perform things like:\n\n> \u201cCompare our GSC queries against competitor keywords and find keywords where competitors rank but we don't, prioritizing low-difficulty terms with meaningful traffic potential.\u201d\n\n### 3. Semrush MCP \u2014 choose this instead of Ahrefs if you're already a Semrush user\n\nSemrush now has an official MCP connection and can be added directly to Claude Code:\n\n```bash\nclaude mcp add semrush https://mcp.semrush.com/v1/mcp -t http\n```\n\nIts MCP exposes organic/paid keywords, keyword research, SERP features, intent, backlinks, competitor analysis and more. [Semrush](https://www.semrush.com/blog/claude-code-seo/?utm_source=chatgpt.com)\n\nI'd choose:\n\n- **Ahrefs** \u2192 if backlinks/content/keyword research are the priority.\n- **Semrush** \u2192 if you already live in Semrush or want broader competitive/market intelligence.\n\nI wouldn't pay for both initially. There's a lot of overlap.\n\n### 4. DataForSEO \u2014 best if you're building an actual SEO agent\n\n[DataForSEO](https://dataforseo.com/?utm_source=chatgpt.com) is more developer-oriented.\n\nIt's particularly attractive if you want Claude Code to **automate SEO research at scale**, rather than simply ask questions of an SEO database.\n\nThink:\n\n```text\nGSC\n \u2193\nidentify opportunities\n \u2193\nDataForSEO\n \u2193\nretrieve SERPs / keyword data\n \u2193\nClaude analyzes competitors\n \u2193\nClaude modifies website/content\n \u2193\nGSC monitors results\n```\n\nDataForSEO's MCP exposes keyword, SERP, backlink and other SEO datasets, and is generally more pay-as-you-go/API-oriented than the big all-in-one SEO subscriptions. [seoskills.sh](https://seoskills.sh/docs/seo-mcp-servers?utm_source=chatgpt.com)\n\n## What I'd actually install\n\nFor a serious Claude Code SEO setup, I'd start with:\n\n```text\nClaude Code\n \u2502\n \u251c\u2500\u2500 Google Search Console MCP\n \u2502 \u2514\u2500\u2500 your actual search performance\n \u2502\n \u2514\u2500\u2500 Ahrefs MCP\n \u251c\u2500\u2500 keyword research\n \u251c\u2500\u2500 competitors\n \u251c\u2500\u2500 backlinks\n \u2514\u2500\u2500 traffic potential\n```\n\nThen optionally add **DataForSEO** if you want Claude to become a more autonomous SEO research/agent system.\n\nThere's also a nice emerging ecosystem of **Claude Code SEO skills** that sit above these MCPs\u2014for example, workflows for weekly GSC reviews, cannibalization detection, indexing audits and content opportunities. [GitHub](https://github.com/acamolese/google-search-console-mcp?utm_source=chatgpt.com)\n\n### The important distinction\n\nI wouldn't think of this as simply *\u201cconnecting Search Console to Claude.\u201d*\n\nThe really valuable setup is:\n\n**First-party data + third-party keyword intelligence + Claude Code skills**\n\nThat gives Claude enough information to make decisions rather than just summarize reports.\n\nFor example, you can eventually have a command like:\n\n```text\n/analyze-seo\n```\n\nthat automatically:\n\n1. Pulls the last 28 days from GSC.\n2. Finds declining pages.\n3. Finds high-impression/low-CTR queries.\n4. Finds position 4\u201320 opportunities.\n5. Cross-references them against Ahrefs/Semrush.\n6. Checks competing pages.\n7. Identifies cannibalization.\n8. Reads the relevant files in your codebase.\n9. Proposes title/meta/content/internal-link changes.\n10. Creates the actual changes as a git diff for you to review.\n\n**If you tell me whether you currently have Ahrefs, Semrush, neither, or a limited budget, I can give you the exact Claude Code MCP setup\u2014including the commands/config and the SEO skills I'd install.**", "sources": [ "https://github.com/LukeRenton/google-search-console-mcp", "https://ahrefs.com/mcp/", "https://ahrefs.com/seo-mcp", "https://www.semrush.com/blog/claude-code-seo/", "https://dataforseo.com/", "https://seoskills.sh/docs/seo-mcp-servers", "https://mcp.directory/blog/best-seo-mcp-servers-2026", "https://github.com/judicael-s/google-search-console-skill", "https://github.com/acamolese/google-search-console-mcp" ], "model": null, "provider_timestamp": null, "web_search": null, "search_queries": [], "served_by": null, "n_sources": 9, "grade": null, "brands": [ "DataForSEO", "Ahrefs", "Semrush", "Search Console" ] }, { "id": "P26", "route": "chatgpt-web-nosearch", "latency_ms": 32456, "fetched_at": "2026-10-07T22:38:44Z", "answer": "If your goal is **programmatic competitor research across both Meta and Google ads**, the key distinction is that neither platform gives you a clean, unrestricted commercial-ad API in the US. Meta\u2019s official API is substantially limited, while Google\u2019s Ads Transparency Center does not offer a general public API for commercial ads. [Facebook](https://br-fr.facebook.com/ads/library/api/?source=onboarding&utm_source=chatgpt.com)\n\n### My shortlist\n\n| API / provider | Meta | Google | Historical data | Best for |\n|---|---|---|---|---|\n| **Foreplay API** | \u2705 | \u2705/varies by product | \u2705 | Best overall for creative intelligence |\n| **Waldo** | \u2705 | \u2705 | \u2705 | Cross-platform monitoring/API workflows |\n| **AdLibrary.com API** | \u2705 | \u2705 | \u2705 | Broad ad-intelligence database |\n| **Apify** | \u2705 | \u2705 | Depends on actor | DIY/custom scraping pipelines |\n| **Meta Ad Library API** | \u26a0\ufe0f | \u2014 | Limited | Official Meta data, especially EU/political |\n| **Google Ads Transparency Center / BigQuery** | \u2014 | \u26a0\ufe0f | Limited | Official Google transparency research |\n\n### 1. Foreplay \u2014 my first choice for creative research\n\n[Foreplay API](https://www.foreplay.co/api?utm_source=chatgpt.com)\n\nForeplay is particularly interesting if you're building a **competitor creative research system**, rather than simply needing raw ad records. Its API advertises access to 300,000+ brands and includes videos/images, creative velocity, transcription, active status, and historical data. It also has a Discovery API for searching its larger ad database. [Foreplay](https://www.foreplay.co/api?utm_source=chatgpt.com)\n\n**Best when you want:**\n- Competitor \u2192 all known ads\n- Creative screenshots/video\n- Ad copy + transcription\n- First/last seen\n- Creative velocity\n- AI analysis of creative patterns\n- An API feeding an internal research agent\n\nI'd put this at **#1 for a marketing/creative-intelligence product**.\n\n### 2. Waldo \u2014 interesting if you're building a cross-platform research agent\n\n[Waldo ad-library API](https://www.waldo.fyi/use-cases/ad-library-api?utm_source=chatgpt.com)\n\nWaldo positions itself specifically around a unified API/MCP layer over **Meta, Google, LinkedIn and TikTok**, with continuously collected historical data. [Waldo](https://www.waldo.fyi/use-cases/ad-library-api?utm_source=chatgpt.com)\n\nThat's compelling if your architecture looks like:\n\n> competitor \u2192 discover ads \u2192 normalize \u2192 analyze \u2192 alert me when something changes\n\nrather than:\n\n> give me a dashboard where I manually browse ads.\n\n### 3. AdLibrary.com \u2014 good broad commercial data layer\n\nThis is worth evaluating if you want **many ad networks through one API** rather than building individual integrations. Its API positioning is explicitly around aggregated ad-library data and competitive intelligence. [adlibrary.com](https://adlibrary.com/posts/ad-spy-api-comparison?utm_source=chatgpt.com)\n\nI'd test it against Foreplay/Waldo on a fixed set of 20\u201350 competitors before committing, because **database freshness and historical depth matter much more than the advertised number of ads**.\n\n### 4. Apify \u2014 best DIY option\n\n[Apify](https://apify.com/?utm_source=chatgpt.com)\n\nIf you're comfortable owning the extraction layer, Apify can be attractive because you can run actors for Meta/Google and turn their datasets into your own normalized database.\n\nThe downside is important: this is fundamentally a **scraping architecture**, not the same thing as having a stable native advertising-data API. Scrapers can break when the underlying site changes, and you need to review each platform's terms before deploying one commercially. [adlibrary.com](https://adlibrary.com/posts/ad-spy-api-comparison?utm_source=chatgpt.com)\n\nI'd choose this if **cost/control > turnkey reliability**.\n\n---\n\n## Don't overlook the official APIs\n\n### Meta\n\n[Meta Ad Library API](https://www.facebook.com/ads/library/api/?utm_source=chatgpt.com)\n\nMeta's official API is excellent\u2014but **not a general-purpose \"give me every competitor's commercial ads\" API**.\n\nCurrently, it provides:\n- Political/social-issue ads globally\n- Ads of any type delivered to the UK/EU during the past year\n- Creative content\n- Page/advertiser information\n- Delivery dates\n- Platforms\n- Spend/impression ranges for political/issue ads\n\nFor ordinary US commercial competitor ads, the API isn't sufficient. [Facebook](https://br-fr.facebook.com/ads/library/api/?source=onboarding&utm_source=chatgpt.com)\n\nThat's why commercial intelligence vendors exist on top of the public library.\n\n### Google\n\nThis is even more important to understand.\n\nThe **Google Ads API is not a competitor-ad-research API**. It's primarily for interacting with *your own Google Ads accounts*. [Google Help](https://support.google.com/google-ads/answer/15235?utm_source=chatgpt.com)\n\nGoogle's **Ads Transparency Center** lets you research verified advertisers' ads, including Search, Display and YouTube, but Google doesn't provide a general public REST API for commercial-ad research. [adlibrary.com](https://adlibrary.com/posts/google-ads-transparency-center-api?utm_source=chatgpt.com)\n\nThere is an official Google transparency dataset in BigQuery, but its scope is limited compared with what you'd ideally want for broad commercial competitive intelligence. [Monocrawl](https://www.monocrawl.com/compare/ad-library-apis?utm_source=chatgpt.com)\n\n---\n\n## What I'd build\n\nIf you're building this for **serious competitor intelligence**, I wouldn't make \"API coverage\" the only selection criterion.\n\nI'd want this normalized schema:\n\n```text\ncompetitor\n \u251c\u2500\u2500 advertiser\n \u251c\u2500\u2500 domain\n \u251c\u2500\u2500 platform\n \u251c\u2500\u2500 ad_id\n \u251c\u2500\u2500 first_seen\n \u251c\u2500\u2500 last_seen\n \u251c\u2500\u2500 active\n \u251c\u2500\u2500 creative_type\n \u251c\u2500\u2500 image/video\n \u251c\u2500\u2500 primary_copy\n \u251c\u2500\u2500 headline\n \u251c\u2500\u2500 CTA\n \u251c\u2500\u2500 landing_page\n \u251c\u2500\u2500 countries\n \u251c\u2500\u2500 placements\n \u251c\u2500\u2500 impressions/reach (when available)\n \u2514\u2500\u2500 source_url\n```\n\nThen add your own derived signals:\n\n- **Creative longevity** \u2014 days/weeks an ad stays active\n- **Creative velocity** \u2014 new ads per week\n- **Offer frequency**\n- **Hook clusters**\n- **CTA clusters**\n- **Landing-page changes**\n- **Advertiser spend/reach proxies**\n- **Winning creative patterns**\n- **Platform overlap**\n- **New competitor launches**\n\nThat turns an ad library into an actual **competitive-intelligence system** rather than another ad-spy dashboard.\n\n### My ranking\n\n**For a new product/team:**\n\n1. **Foreplay** \u2014 best creative intelligence/API starting point\n2. **Waldo** \u2014 strongest concept for unified cross-platform programmatic research\n3. **AdLibrary.com** \u2014 broad data/API option worth benchmarking\n4. **Apify** \u2014 best if you want to own the extraction infrastructure\n5. **Official Meta/Google APIs** \u2014 use wherever their coverage actually matches your geography/use case\n\nThe most important thing I'd do before buying anything is run a **50-competitor bake-off**: 25 Meta-heavy brands + 25 Google-heavy brands, then compare *recall, freshness, historical depth, creative assets, and API latency*. The vendors' \"millions of ads\" claims aren't nearly as useful as whether they consistently find the ads you actually care about. [adlibrary.com](https://adlibrary.com/posts/ad-spy-api-comparison?utm_source=chatgpt.com)", "sources": [ "https://br-fr.facebook.com/ads/library/api/?source=onboarding", "https://adlibrary.com/posts/google-ads-transparency-center-api", "https://www.foreplay.co/api", "https://www.waldo.fyi/use-cases/ad-library-api", "https://adlibrary.com/posts/ad-spy-api-comparison", "https://apify.com/", "https://www.facebook.com/ads/library/api/", "https://support.google.com/google-ads/answer/15235", "https://www.monocrawl.com/compare/ad-library-apis" ], "model": null, "provider_timestamp": null, "web_search": null, "search_queries": [], "served_by": null, "n_sources": 9, "grade": null, "brands": [ "Apify", "Meta Ad Library", "Ads Transparency", "Foreplay" ] }, { "id": "F1", "route": "gemini", "latency_ms": 43353, "fetched_at": "2026-10-07T22:34:49Z", "answer": "The multimodal filter in Google Search Console is a feature within the Performance report that allows site owners to track and isolate web traffic coming from image-assisted and visual searches.\n\nWhat the Multimodal Filter Covers\n\nWhen you select the Web search type in your Performance report, Google Search Console provides an option to split traffic into text-based searches and multimodal searches. The multimodal filter tracks web result journeys that began with an image instead of just text, including:\n\nGoogle Lens searches\n\nCircle to Search on Android\n\nImages or screenshots uploaded directly to Google Search\n\nChrome's \"Search this image\" right-click feature\n\nKey Features and Limitations\n\nPage-Level Insights: You can view clicks, impressions, click-through rates (CTR), and device/country breakdowns to see which of your landing pages (such as product pages, local listings, or documentation) are being discovered via visual search.\n\nNo Query Data: Because these searches are driven by images rather than text strings, specific text query data is unavailable when this filter is applied. The Queries tab is hidden for multimodal views, meaning you cannot see the exact photo or text prompt a user submitted.\n\nNot Separated from Traditional Image Search: This filter isolates web results that were triggered by visual discovery (e.g., someone taking a picture of a product part and landing on an e-commerce or help page), rather than traditional Google Images tab clicks.\n\nWhy It Matters for SEO\n\nWith visual discovery tools like Google Lens handling billions of queries, this filter gives marketers and site owners a way to measure how effectively their visual assets (such as clear product photography, diagrams, or screenshots) are driving real-world discovery and traffic.", "sources": [ "https://almcorp.com/news/search-console-adds-multimodal-filter-image-searches/#:~:text=Google%20is%20adding%20a%20new%20multi%2Dmodal%20filter,Central%20blog%20and%20is%20rolling%20out%20globally.", "https://suganthan.com/notes/search-console-multimodal/#:~:text=Search%20Console%20now%20reports%20multimodal%20search.%20You,not%20the%20query%20that%20prompted%20the%20search.", "https://nugentive.com/blog/search-console-multimodal-filter/#:~:text=Google's%20documentation%20lists%20Web%3A%20multimodal%20as%20web,give%20the%20visitor%20an%20obvious%20next%20step.", "https://almcorp.com/news/search-console-adds-multimodal-filter-image-searches/#:~:text=Google%20is%20adding%20a%20new%20multi%2Dmodal%20filter,Central%20blog%20and%20is%20rolling%20out%20globally.", "https://suganthan.com/notes/search-console-multimodal/#:~:text=You%20can%20see%20the%20pages%2C%20clicks%20and,distinction%20on%20its%20Performance%20report%20help%20page.", "https://nugentive.com/blog/search-console-multimodal-filter/#:~:text=According%20to%20Google%2C%20the%20reporting%20includes%20searches,*%20Chrome's%20right%2Dclick%20%E2%80%9CSearch%20this%20image%E2%80%9D%20feature.", "https://nugentive.com/blog/search-console-multimodal-filter/#:~:text=Owners%20can%20still%20analyze%20landing%20pages%2C%20clicks%2C,ordinary%20product%2C%20service%2C%20guide%2C%20or%20location%20page.", "https://www.searchenginejournal.com/google-search-console-multimodal-filter/590781/#:~:text=Multimodal%20traffic%20comes%20without%20query%20data.%20Google's,to%20the%20Export%20button%20in%20the%20report.", "https://www.searchenginejournal.com/google-search-console-multimodal-filter/590781/#:~:text=Multimodal%20traffic%20comes%20without%20query%20data.%20Google's,to%20the%20Export%20button%20in%20the%20report.", "https://nugentive.com/blog/search-console-multimodal-filter/#:~:text=Owners%20can%20still%20analyze%20landing%20pages%2C%20clicks%2C,ordinary%20product%2C%20service%2C%20guide%2C%20or%20location%20page.", "https://suganthan.com/notes/search-console-multimodal/#:~:text=You%20can%20see%20the%20pages%2C%20clicks%20and,distinction%20on%20its%20Performance%20report%20help%20page." ], "model": null, "provider_timestamp": null, "web_search": null, "search_queries": null, "served_by": null, "n_sources": 11, "grade": "correct", "brands": [ "Search Console" ] }, { "id": "F2", "route": "gemini", "latency_ms": 42302, "fetched_at": "2026-10-07T22:35:31Z", "answer": "Ahrefs did not introduce major structural changes to its overall API or Connect unit limits in September 2026. However, a few notable API updates, deprecations, and parameter modifications went into effect during that month:\n\nWeb Analytics API: The unknown value was removed from the source_channel enum on Web Analytics API endpoints on September 15, 2026, with any leftover requests filtering on unknown defaulting to referral.\n\nSite Explorer API: The default value for traffic_mode shifted from static to adaptive on September 4, 2026.\n\nBrand Radar API: The grace period for the old search_volume_type=keyword_volume option sunsetted on September 30, 2026, transitioning all requests to AI-adjusted volume.\n\nNew Endpoints: A new Content Helper Topics endpoint was launched to score target URLs against top-ranking pages.\n\n(Note: If you are thinking of broader quota expansions\u2014such as the massive multi-fold increases to Ahrefs Connect units, credit limits, and MCP/Direct API row caps\u2014those actually rolled out slightly earlier in the summer).", "sources": [ "https://docs.ahrefs.com/en/api/docs/changelog#:~:text=Changelog%20*%20Sep%2015%2C%202026.%20The%20unknown,has%20changed%20from%20static%20to%20adaptive%20.", "https://docs.ahrefs.com/en/api/docs/changelog#:~:text=*%20Sep%204%2C%202026.%20A%20new%20Content,using%20the%20previous%20organic%20traffic%20calculation%20model.", "https://docs.ahrefs.com/en/api/docs/changelog#:~:text=The%20default%20value%20of%20traffic_mode%20on%20Site,endpoint%20now%20returns%20each%20prompt's%20tags%20field.", "https://docs.ahrefs.com/en/api/docs/changelog#:~:text=*%20Sep%204%2C%202026.%20A%20new%20Content,using%20the%20previous%20organic%20traffic%20calculation%20model.", "https://pricingsaas.com/companies/ahrefs/diffs/2026Q2#:~:text=Before%20and%20after.%20Mar%2030%2C%202026%20vs,units%20Lite%2025%2C000%E2%86%92100%2C000%2C%20Standard%20150%2C000%E2%86%92400%2C000%2C%20Advanced%20500%2C000%E2%86%921M%3B" ], "model": null, "provider_timestamp": null, "web_search": null, "search_queries": null, "served_by": null, "n_sources": 5, "grade": "wrong", "brands": [ "Ahrefs" ] }, { "id": "F3", "route": "gemini", "latency_ms": 43650, "fetched_at": "2026-10-07T22:36:15Z", "answer": "OpenSEO is an open-source, self-hostable alternative to traditional SEO suites like Semrush and Ahrefs. It provides a full set of search engine optimization features along with native support for AI assistants via the Model Context Protocol (MCP).\n\nKey Features\n\nAI Agent Integration (MCP Server): Allows AI clients like Claude, Cursor, ChatGPT, and Codex to access live SERPs, inspect keyword metrics, pull backlink profiles, and query Search Console data directly within prompt sessions.\n\nCore SEO Suite:\n\nKeyword Research & Rank Tracking: Tracks keyword volume, difficulty, search intent, CPC, and historical positions over time.\n\nDomain & Backlink Analysis: Evaluates estimated organic traffic, top pages, target URLs, referring domains, and backlink health.\n\nTechnical Site Auditing: Features a built-in crawler to analyze titles, status codes, meta tags, indexability, and site health.\n\nAI Visibility Monitoring: Tracks brand mentions, prompts, and citations across AI search engines like ChatGPT and Google AI Overviews.\n\nDeployment Options: Can be self-hosted via Docker/Cloudflare (using your own API keys) or accessed as a hosted web app (at openseo.so).\n\nData Sources\n\nOpenSEO draws its data from a combination of third-party APIs and built-in systems:\n\nData Type\nPrimary Source\nDetails\nSEO & SERP Intelligence\nDataForSEO\nPowers keyword volume, live SERP results, competitor traffic metrics, backlink profiles, and difficulty scores.\nFirst-Party Analytics\nGoogle Search Console\nConnects directly via official Google APIs for real-time site clicks, impressions, CTR, position, and URL inspection data.\nSite Crawling & Audits\nCustom Engine & DataForSEO\nUses OpenSEO's native custom crawler for site structure/page-level issues, supplemented by DataForSEO for optional page-speed/Lighthouse audits.", "sources": [ "https://ryandoser.com/openseo-review/#:~:text=GitHub%20lists%2019%20contributors%2C%20but%20Ben%20Senescu,%22No%20subscriptions%22%20only%20applies%20if%20you%20self%2Dhost.", "https://ryandoser.com/openseo-review/#:~:text=It%20covers%20keyword%20research%2C%20rank%20tracking%2C%20backlinks%2C,DataForSEO.%20The%20hosted%20version%20is%20at%20openseo.so.", "https://www.microsaasexamples.com/p/openseo#:~:text=Credits%20are%20used%20by%20DataForSEO%20API%20features,Google%20Search%20Console%20access%20read%20only%3F%20Yes.", "https://ryandoser.com/openseo-review/#:~:text=Its%20GitHub%20repository%20passed%2020%2C000%20stars%20in,answers%20%22where%20do%20I%20rank%20right%20now%3F%22", "https://www.microsaasexamples.com/p/openseo#:~:text=OpenSEO%20FAQ.%20What%20does%20OpenSEO%20do%3F%20OpenSEO,managed%20app%20starts%20at%20%2410%20per%20month.", "https://ryandoser.com/openseo-review/#:~:text=Several%20pieces%20are%20OpenSEO's%20own%20work%3A%20*,agent%20reads%20them%20when%20it%20starts%20work.", "https://openseo.so/features#:~:text=Explore%20feature%20%E2%86%92%20SEO%20audit%20tool%20for,visibility%20research.%20Compare%20answers%20across%20supported%20models.", "https://ryandoser.com/openseo-review/#:~:text=It%20covers%20keyword%20research%2C%20rank%20tracking%2C%20backlinks%2C,DataForSEO.%20The%20hosted%20version%20is%20at%20openseo.so." ], "model": null, "provider_timestamp": null, "web_search": null, "search_queries": null, "served_by": null, "n_sources": 8, "grade": "correct", "brands": [ "DataForSEO", "Ahrefs", "Semrush", "Search Console" ] }, { "id": "F4", "route": "gemini", "latency_ms": 38773, "fetched_at": "2026-10-07T22:36:54Z", "answer": "Jev is a proprietary artificial intelligence model developed by TypeSafe AI (founded by former OpenAI researcher Diogo Almeida). Released in September 2026, Jev represents a distinct class of AI models that the company calls a \"System One model\"\u2014named after Daniel Kahneman's concept of fast, intuitive, subconscious thinking.\n\nUnlike traditional generative large language models (LLMs) like GPT-4 or Claude, which output natural language token-by-token, Jev is built specifically to make fast, structured decisions for software systems.\n\nCore Characteristics of Jev\n\nNo Text Generation: Jev does not write prose, generate code, summarize text, or explain its reasoning. It evaluates a piece of input text or JSON (called the state) against a set of typed questions defined in code, and returns strictly typed values, probabilities, and confidence scores.\n\nStructured Output Guarantee: Because its outputs are schema-compliant and bound to predefined choices, it structurally eliminates parsing failures, formatting errors, or markdown-fence issues common when forcing LLMs to output JSON.\n\nSupported Question Types: Every API call is built from three native question types:\n\nChoice: Picks one option from an unordered list of up to 255 items, returning probabilities and a confidence score.\n\nScore: Places the input on an ordered scale of 2 to 10 levels described in words.\n\nNoul: A binary (yes/no) proposition that returns a single probability between 0 and 1.\n\nSpeed and Cost: Because it evaluates states and questions in a single parallel pass rather than autoregressively generating strings, Jev operates with end-to-end response times between 70 to 500 milliseconds, making it significantly faster and cheaper than running classification tasks through a frontier LLM.\n\nWhat Problem Does It Solve?\n\nIn many software pipelines, developers use heavy generative LLMs just to route a task\u2014such as classifying a support ticket, checking content moderation, extracting fields, or determining a workflow branch. Traditionally, this requires writing fragile wrapper code to force the LLM to format its response as JSON, parse it, and handle retries when it fails.\n\nJev bypasses this entire overhead by acting as a native, probabilistic decision function meant to be consumed directly by software logic rather than read by humans.", "sources": [ "https://en.wikipedia.org/wiki/Jev_(AI_model)#:~:text=Jev%20is%20a%20proprietary%20artificial%20intelligence%20model,led%20by%20DCVC.%20Jev.%20Developer.%20TypeSafe%20AI.", "https://www.sanity.io/glossary/jev-typesafe-ai-model#:~:text=TypeSafe%20AI%20released%20Jev%20on%2015%20September,and%20a%20co%2Dauthor%20of%20the%20InstructGPT%20paper.", "https://medium.com/@unicodeveloper/the-ultimate-guide-to-jev-the-new-frontier-ai-for-faster-decisions-acd78e5f4c56#:~:text=A%20System%20One%20model%20is%20a%20class,been%20paying%20System%202%20prices%20for%20them.", "https://www.sanity.io/glossary/jev-typesafe-ai-model#:~:text=There%20is%20no%20text%20output%20of%20any,Generative%20models%20do%20the%20deliberate%2C%20verbose%20thing.", "https://www.sanity.io/glossary/jev-typesafe-ai-model#:~:text=Jev%20(TypeSafe%20AI%20model)%20definition%20*%20What,state%2C%20write%20a%20rationale%2C%20or%20produce%20code.", "https://www.sanity.io/glossary/jev-typesafe-ai-model#:~:text=Jev's%20output%20is%20constrained%20to%20schema%2Dvalid%20values%2C,never%20fail%20is%20useful%20in%20a%20pipeline.", "https://www.sanity.io/glossary/jev-typesafe-ai-model#:~:text=The%20implication%20TypeSafe%20is%20drawing%20is%20that,won%20and%20how%20close%20the%20runners%2Dup%20were.", "https://www.sanity.io/glossary/jev-typesafe-ai-model#:~:text=The%20implication%20TypeSafe%20is%20drawing%20is%20that,won%20and%20how%20close%20the%20runners%2Dup%20were.", "https://www.sanity.io/glossary/jev-typesafe-ai-model#:~:text=Choice%20picks%20one%20option%20from%20a%20list,than%20being%20forced%20onto%20a%20single%20rung.", "https://www.sanity.io/glossary/jev-typesafe-ai-model#:~:text=Noul%20returns%20a%20single%20probability%20between%200,claim%20is%20about%20format%2C%20not%20about%20correctness.", "https://medium.com/@unicodeveloper/the-ultimate-guide-to-jev-the-new-frontier-ai-for-faster-decisions-acd78e5f4c56#:~:text=An%20LLM%20produces%20its%20answer%20one%20token,an%20answer%20space%20you%20defined%20in%20advance.", "https://medium.com/@unicodeveloper/the-ultimate-guide-to-jev-the-new-frontier-ai-for-faster-decisions-acd78e5f4c56#:~:text=What%20Exactly%20does%20Jev%20Solve%3F%20*%20What,as%20%22technical%20support%22%20instead%20of%20technical%20.", "https://en.wikipedia.org/wiki/Jev_(AI_model)#:~:text=It%20does%20not%20generate%20natural%2Dlanguage%20text.%20It,models%2C%20Jev%20does%20not%20require%20data%2Dspecific%20training." ], "model": null, "provider_timestamp": null, "web_search": null, "search_queries": null, "served_by": null, "n_sources": 13, "grade": "correct", "brands": [] }, { "id": "P02", "route": "gemini", "latency_ms": 37880, "fetched_at": "2026-10-07T22:40:17Z", "answer": "If you want to enrich company domains (pulling firmographics, tech stacks, employee count, industry, etc.) without committing to a recurring monthly subscription, you have a few cost-effective, non-subscription options.\n\nThe cheapest and most practical paths depend on whether you prefer pay-as-you-go credit systems, free tiers, or DIY programmatic scraping.\n\n1. Pay-As-You-Go (No Subscription) Credit APIs\n\nInstead of paying a monthly fee, several data providers and APIs let you buy one-time credit bundles that never expire or only charge per successful lookup.\n\nOutscraper / Google Maps & Custom Scrapers: Excellent if you want to extract firm-level data linked to domains or business profiles pay-as-you-go style. Costs fractions of a cent per row.\n\nDropcontact / Proxycurl (API-based): Offer pay-as-you-go top-ups where you only pay for the exact volume of domains you enrich, avoiding subscription lock-in.\n\nEnrichmentAPI.io / CompanyEnrich: Provide developer-focused REST APIs with usage-based or pay-as-you-go tiers for clean JSON firmographic data.\n\n2. Generous Free Tiers & Freemium Platforms\n\nIf your volume is low to medium, you can cycle through or utilize freemium tiers without ever upgrading to a paid monthly plan:\n\nApollo.io: Offers a robust free tier with generous monthly search and company enrichment credits. You can look up firmographics, industries, and employee sizes directly from domains.\n\nHunter.io: Provides free monthly domain searches that reveal email patterns, employee listings, and basic company association.\n\nSyncGTM / Click5 Enrich / Oppora: Newer GTM and enrichment platforms that offer free tiers or trial credits specifically designed to test automated company data appending.\n\n3. The DIY Route: Open-Source Code & Custom Scrapers (Absolute Cheapest)\n\nIf you have a bit of technical know-how (or use an AI assistant to write the script), you can enrich domains yourself for pennies (cost of proxy/server time only):\n\nDNS and SSL Certificate Inspection: Write a short Python script using libraries like requests and ssl to look up a domain's DNS records, MX records, and SSL organization details. SSL certs often reveal the legal entity name behind a domain for free.\n\nMeta-Tag & Footer Scraper: Write a lightweight web scraper (using Python's BeautifulSoup or Cheerio in Node.js) to visit the homepages of your list of domains. It can automatically extract:\n\nCopyright footers (often containing the exact legal company name).\n\nOpenGraph meta tags and social links (LinkedIn company page, Twitter/X, GitHub).\n\nTechnographic footprints (looking for tracking codes like Google Analytics, Shopify, WordPress, or HubSpot in the source code).\n\nOpen-Source Databases: Cross-reference your domain list against open datasets or public registries (like Crunchbase data dumps or open government business registries) locally via CSV or a local SQL database.\n\nSummary Recommendation\n\nLowest Effort, Zero Cost: Use Apollo.io's free tier or free API credits to batch-upload or search your domains.\n\nTrue Pay-As-You-Go: Use an API like Proxycurl or Outscraper to buy a one-time bucket of credits without a recurring subscription.\n\nAt Scale / Tech-Savvy: Spin up a custom Python script to scrape homepage meta-tags and check SSL/DNS records for virtually $0.", "sources": [ "https://click5enrich.com/free-data-enrichment-api/#:~:text=Overview.%20EnrichmentAPI.io%20focuses%20on%20simple%20and%20fast,lookups%20and%20returns%20clean%20JSON%20data.%20Highlights%3A", "https://syncgtm.com/blog/best-free-enrichment-tools-2026#:~:text=Apollo.io%20has%20the%20most%20generous%20free%20tier,credit%20limits%20being%20the%20only%20real%20constraint.", "https://syncgtm.com/blog/best-free-enrichment-tools-2026#:~:text=Organization%20%24119%2Fmo.%20*%203.%20Hunter.io%20%E2%80%94%20Email,company%2C%20revealing%20email%20patterns%20and%20individual%20contacts." ], "model": null, "provider_timestamp": null, "web_search": null, "search_queries": null, "served_by": null, "n_sources": 3, "grade": null, "brands": [ "Apollo", "Hunter", "Crunchbase" ] }, { "id": "P21", "route": "gemini", "latency_ms": 40795, "fetched_at": "2026-10-07T22:37:34Z", "answer": "For an in-house growth agent or data engineering team building automated SEO pipelines, workflows, or AI agents, the ideal API stack depends on whether you need raw search/SERP extraction, massive backlink and keyword databases, or AI-driven search visibility (GEO).\n\nThe top SEO data APIs for an in-house growth stack are categorized below by their core strengths:\n\n1. Best for Modular & Developer-First Pipelines: DataForSEO\n\nPrimary Use Case: Comprehensive, multi-engine SEO data (SERP, keyword research, backlinks, on-page audits, and AI/GEO visibility) via a modular pay-as-you-go API.\n\nWhy it fits a growth agent: Unlike traditional SaaS tools that force you into a fixed dashboard subscription, DataForSEO is a pure API platform with no minimum monthly tiers. It covers Google, Bing, YouTube, Amazon, and AI search engines. It also features built-in support for AI agent setups (including Model Context Protocol/MCP servers), making it seamless to hook directly into LLMs and workflow automation tools like n8n or Claude Code.\n\n2. Best for Live AI Agent Workflows & Web Data: Nimble\n\nPrimary Use Case: Live search data, proxy orchestration, structured SERP output, and AI agent intelligence pipelines.\n\nWhy it fits a growth agent: Nimble is built specifically for engineering and growth teams creating AI-native pipelines. It handles complex proxy management, rendering, and geo-targeting under the hood, returning clean JSON data from live SERPs. Its specialized SEO agent modules help ingest live search dynamics straight into data warehouses like Snowflake, Databricks, or S3.\n\n3. Best for Deep Backlink & Competitive Intelligence: Ahrefs API\n\nPrimary Use Case: Enterprise-grade backlink indexes, historical organic performance, and domain authority metrics.\n\nWhy it fits a growth agent: If your growth workflows heavily rely on link-building automation, digital PR tracking, or deep historical competitor analysis, Ahrefs offers the industry standard index. While higher-end enterprise plans can be costly, its machine-readable endpoints provide unmatched granularity on referring domains, anchor texts, and traffic estimations. (Semrush is a strong alternative here if your focus leans more heavily toward PPC and keyword semantic databases).\n\n4. Best for Pure SERP Extraction & Reliability: SerpApi\n\nPrimary Use Case: Real-time structured parsing of Google and multi-engine search results, maps, and news.\n\nWhy it fits a growth agent: When you need bulletproof rank tracking without dealing with CAPTCHAs, proxy blocking, or layout changes, SerpApi acts as a reliable abstraction layer. It turns complex search pages into structured JSON instantly and integrates effortlessly with custom scripts or no-code/low-code growth tech stacks.\n\n5. Best for Emerging AI Search (GEO) Tracking: AIclicks or Cloro\n\nPrimary Use Case: Tracking brand mentions and visibility across generative AI engines (ChatGPT, Claude, Perplexity, Google AI Overviews, Copilot).\n\nWhy it fits a growth agent: Modern growth agents need to look beyond traditional blue links. These specialized APIs track how your brand and competitors are cited inside AI answers, filling the blind spot that traditional SEO tools often miss.\n\nSummary Recommendation for a Growth Stack\n\nFor a flexible, all-in-one developer setup: Start with DataForSEO for core keywords, SERPs, and backlinks on a consumption model.\n\nFor autonomous AI agent execution & live web scraping: Pair it with Nimble or SerpApi.\n\nFor deep link authority metrics: Integrate Ahrefs if budget allows for enterprise-grade intelligence.\n\nWhat specific growth workflows (e.g., automated rank tracking, programmatic content scaling, or backlink prospecting) are you building this agent for?", "sources": [ "https://www.nimbleway.com/blog/top-seo-analysis-apis#:~:text=*%201.%20Nimble%20%E2%80%93%20Recommended%20for%20engineering,orchestration%2C%20extraction%2C%20and%20reliable%20delivery%20at%20scale.", "https://www.position.digital/blog/best-seo-geo-api/#:~:text=*%20Pay%2Das%2Dyou%2Dgo%20pricing.%20Most%20AI%20visibility%20tracking,five%20data%20types.%20*%20AI%20Optimization%20API.", "https://www.nimbleway.com/blog/top-seo-analysis-apis#:~:text=2.%20DataForSEO%20%E2%80%93%20Recommended%20for%20building%20custom,through%20a%20modular%20endpoint%20structure.%20Key%20Features.", "https://www.nimbleway.com/blog/top-seo-analysis-apis#:~:text=Built%20on%20a%20continuously%20updating%20database%20of,the%20comprehensive%20nature%20of%20the%20APIs%20available.%E2%80%9D", "https://www.position.digital/blog/best-seo-geo-api/#:~:text=DataForSEO%20*%20Best%20for%3A%20SEO%20agencies%20and,AI%20Overview%2C%20ChatGPT%2C%20Gemini%2C%20Claude%2C%20and%20Perplexity.", "https://www.nimbleway.com/blog/top-seo-analysis-apis#:~:text=*%201.%20Nimble%20%E2%80%93%20Recommended%20for%20engineering,orchestration%2C%20extraction%2C%20and%20reliable%20delivery%20at%20scale.", "https://www.nimbleway.com/blog/top-seo-analysis-apis#:~:text=*%201.%20Nimble%20%E2%80%93%20Recommended%20for%20engineering,orchestration%2C%20extraction%2C%20and%20reliable%20delivery%20at%20scale.", "https://www.nimbleway.com/blog/top-seo-analysis-apis#:~:text=Together%2C%20Nimble's%20SERP%20API%20and%20SEO%20agent,managed%20infrastructure%20for%20rendering%2C%20access%2C%20and%20scale.", "https://www.nimbleway.com/blog/top-seo-analysis-apis#:~:text=High%2Dthroughput%20scaling%20with%20dedicated%20and%20rotated%20residential,as%20Snowflake%2C%20Databricks%2C%20or%20S3%20where%20configured.", "https://seoprofy.com/blog/seo-data-api/#:~:text=Ahrefs%20*%20Direct%20machine%2Dreadable%20endpoints%20for%20lead,(AWT)%20programmatic%20keys%20for%20verified%20site%20owners.", "https://growwer.com/blog/the-best-seo-apis/#:~:text=*%20Which%20SEO%20API%20is%20best%20for,*%20SerpApi%3A%20structured%20extraction%20of%20search%20results.", "https://seoprofy.com/blog/seo-data-api/#:~:text=Ahrefs%20*%20Direct%20machine%2Dreadable%20endpoints%20for%20lead,(AWT)%20programmatic%20keys%20for%20verified%20site%20owners.", "https://www.position.digital/blog/best-seo-geo-api/#:~:text=AIclicks%20*%20Best%20for%3A%20Teams%20who%20want,API%20access%20on%20the%20%2459%20Starter%20plan.", "https://www.position.digital/blog/best-seo-geo-api/#:~:text=DataForSEO%20does%2C%20at%20a%20fraction%20of%20the,AI%20citations%20come%20from%20the%20same%20provider.", "https://growwer.com/blog/the-best-seo-apis/#:~:text=*%20Use%20cases%20for%20agencies%20and%20developers.,metrics.%20*%20DataForSEO%3A%20modular%20flexibility%20and%20pay%2Das%2Dyou%2Dgo.", "https://www.nimbleway.com/blog/top-seo-analysis-apis#:~:text=Top%20SEO%20Analysis%20APIs%20for%202026%20include%3A,Console%20API.%20*%20Google%20PageSpeed%20Insights%20API." ], "model": null, "provider_timestamp": null, "web_search": null, "search_queries": null, "served_by": null, "n_sources": 16, "grade": null, "brands": [ "DataForSEO", "Ahrefs", "Semrush", "SerpApi" ] }, { "id": "P22", "route": "gemini", "latency_ms": 43286, "fetched_at": "2026-10-07T22:38:18Z", "answer": "Getting keyword volumes and backlinks without paying for multiple high-priced SEO subscriptions is entirely possible by combining free tools, browser extensions, and built-in platform data.\n\nPart 1: How to Get Keyword Volumes for Free\n\nInstead of paying for a dedicated keyword tool, you can leverage direct search data sources and free tiers:\n\nGoogle Keyword Planner: The foundational source for Google search data. While it is designed for Google Ads and typically shows broad search volume ranges unless you run active campaigns, it remains completely free and provides direct-from-the-source data.\n\nFree Standalone Volume Checkers: Tools like SearchVolume.io allow you to bulk-check thousands of keywords at once using cached database estimates modeled after Google Ads data. Similarly, Semrush's Free Keyword Search Volume Checker and WordStream's Free Keyword Tool let you run limited queries to evaluate demand and intent without a credit card.\n\nFreemium Browser Extensions: Tools like Keywords Everywhere can be installed on Chrome or Firefox. While deep historical trend charts and bulk uploads require a cheap, pay-as-you-go credit model (rather than a heavy monthly fee), the extension provides massive amounts of free passive data\u2014including related keywords, \"People Also Ask\" metrics, and traffic insights right inside your browser.\n\nAlternative Discovery (Zero-Cost): Use Google's own Autocomplete suggestions, the People Also Ask boxes on search engine result pages (SERPs), and prompts via AI chatbots to flesh out massive lists of long-tail keywords organically.\n\nPart 2: How to Track and Analyze Backlinks for Free\n\nBacklink data is notoriously guarded by enterprise platforms, but you can bypass subscriptions using native dashboards and free audits:\n\nGoogle Search Console (GSC): This is your most valuable, completely free asset. GSC tells you exactly which external sites are linking to you, your total link count, your top-linked pages, and your internal linking structure.\n\nFree Backlink Checkers (Freemium Tiers): Major platforms offer limited free lookups if you register for a free account. For example, tools like Ahrefs Backlink Checker or Semrush allow you to plug in any domain to see a snapshot of its Domain Authority/Rating, top referring domains, and a sample of active backlinks without starting a paid subscription.\n\nOpen-Source and Community Scrapers: If you have technical skills or want to crawl smaller sites, tools like Screaming Frog SEO Spider (free up to 500 URLs) let you audit internal links and site architecture locally on your computer.\n\nPart 3: A Practical Zero-to-Low-Cost Workflow\n\nFor Keyword Research: Brainstorm seed terms, drop them into SearchVolume.io or WordStream's Free Keyword Tool to map out demand, and use Keywords Everywhere for quick SERP validation.\n\nFor Competitor Backlinks: Paste your competitor\u2019s URL into a free backlink checker lookup to see who links to them, then manually target those same sites (e.g., resource pages, guest post directories, or industry roundups) via outreach.\n\nFor Performance Tracking: Monitor your actual keyword rankings, impressions, clicks, and incoming backlink profile exclusively through Google Search Console.\n\nWhat specific niche or type of website are you gathering keywords for?", "sources": [ "https://onelittleweb.com/top-tools/best-keyword-research-tools/#:~:text=Google%20Keyword%20Planner.%20Bootstrapped%20startups%20with%20zero,filtering%20but%20data%20straight%20from%20Google.%20Ubersuggest.", "https://keywordseverywhere.com/#:~:text=From%20install%20to%20insights%20in%20a%20couple,search%20volume%2C%20CPC%2C%20competition%20on%2015%2B%20sites.", "https://www.wordstream.com/keywords#:~:text=Other%20free%2C%20simple%20ways%20to%20uncover%20long,queries%2C%20and%20asking%20AI%20chatbots%20for%20suggestions.", "https://keywordseverywhere.com/#:~:text=From%20install%20to%20insights%20in%20a%20couple,search%20volume%2C%20CPC%2C%20competition%20on%2015%2B%20sites." ], "model": null, "provider_timestamp": null, "web_search": null, "search_queries": null, "served_by": null, "n_sources": 4, "grade": null, "brands": [ "Ahrefs", "Semrush", "Search Console", "Keywords Everywhere" ] }, { "id": "P23", "route": "gemini", "latency_ms": 40750, "fetched_at": "2026-10-07T22:38:59Z", "answer": "Running weekly keyword and SERP checks cost-effectively through an API depends heavily on whether you need raw SERP rank tracking (where your page ranks for specific keywords) or keyword metrics (search volume, CPC, and difficulty).\n\nEnterprise tools like Semrush, Ahrefs, or SerpApi lock you into expensive monthly subscriptions ($25 to $150+/month). To build a budget-friendly pipeline, use the following strategies:\n\n1. For SERP & Rank Tracking (Cheapest APIs)\n\nIf you want to check what URLs rank in Google for a given set of keywords on a weekly basis, pay-as-you-go providers with no monthly minimums offer the lowest cost:\n\nSerpent API: Highly cost-effective for raw SERP data, with rates starting as low as $0.03 to $0.60 per 1,000 searches depending on volume tiers, with zero monthly minimums.\n\nDataForSEO: A favorite for developers because of its true pay-as-you-go model (with a small $50 minimum deposit). SERP checks cost roughly $0.60 per 1,000 requests, and you only pay for what you query.\n\nSearlo / PrismCrawl: Offer prepaid credit packs starting around $0.19 to $0.30 per 1,000 searches for high-volume users, though credits typically expire after 90 days.\n\n2. For Keyword Volume and Metrics (Cheapest APIs)\n\nIf you need search volumes, competition metrics, or keyword ideas alongside your checks:\n\nDataForSEO (Google Ads API Endpoint): Easily the cheapest bulk source for keyword metrics. It charges roughly $0.06 per task (which can bundle up to 1,000 keywords in the standard queue), translating to about $60 per million keywords.\n\nGoogle Search Console (GSC) API (100% Free): If you only care about keywords your site already ranks for, the GSC API is completely free and has no row limits or costs. You can pull impressions, clicks, and average positions programmatically as often as you like.\n\n3. The Ultimate DIY Alternative: Serverless + Scraping\n\nIf you have programming experience and want to eliminate third-party API markup entirely, you can host your own checker:\n\nThe Setup: Write a lightweight script using Python (Playwright or BeautifulSoup) or Node.js, and deploy it as a scheduled AWS Lambda function, Cloudflare Worker, or GitHub Action (which offers free CI/minute quotas that easily handle weekly runs).\n\nThe Catch: Google aggressively blocks data centers. To prevent your scraper from getting blocked or returning CAPTCHAs, you will need to route requests through a cheap datacenter or residential proxy rotator (e.g., Bright Data, Oxylabs, or Webshare proxy pools).\n\nSummary Recommendation\n\nLowest Effort & Lowest Cost: Use DataForSEO or Serpent API on a pay-as-you-go basis. For 1,000 keywords checked weekly (4,000 checks a month), your total cost will be less than $3.00 a month with zero fixed subscriptions.\n\nZero Cost (Existing Rankings Only): Script a weekly fetch using the Google Search Console API.\n\nHow many keywords are you planning to track on a weekly basis?", "sources": [ "https://apiserpent.com/blog/serp-api-pricing-comparison#:~:text=What%20pricing%20models%20do%20SERP%20API%20providers,from%20%2425%2Fmonth%20for%201%2C000%20searches%20at%20SerpApi.", "https://apiserpent.com/blog/serp-api-pricing-comparison#:~:text=This%20makes%20subscription%20models%20particularly%20wasteful.%20A,(100%2C000%2B%20searches)%20remain%20within%20typical%20academic%20budgets.", "https://apiserpent.com/blog/serp-api-pricing-comparison#:~:text=Table_title%3A%20Full%20Comparison%20Table%20Table_content%3A%20%7C%20Provider,credits%2Fmonth%20%7C%20No%20(monthly%20credit%20plans)%20%7C", "https://apiserpent.com/blog/serp-api-pricing-comparison#:~:text=Pay%2DPer%2DSearch%20(True%20Pay%2Das%2DYou%2DGo)%20In%20this%20model%2C%20you,pay%20for%20exactly%20what%20you%20use%E2%80%94nothing%20more.", "https://www.prismcrawl.com/blog/best-google-serp-api#:~:text=Evaluation%20consideration.%20PrismCrawl.%20One%2Dtime%20prepaid%20credits.%20As,requests.%20PAYG%20is%20an%20option%20here%20too.", "https://answersocrates.com/blog/best-keyword-research-apis/#:~:text=It%20depends%20entirely%20on%20the%20billing%20unit.,wildly%20different%20bills%20for%20the%20same%20workload.", "https://answersocrates.com/blog/best-keyword-research-apis/#:~:text=Skip%20it%20if%20you%20want%20a%20dashboard,list%20tells%20you%20what%20the%20world%20searches.", "https://answersocrates.com/blog/best-keyword-research-apis/#:~:text=There%20is%20no%20official%20keyword%20volume%20API,undervalued%20free%20data%20source%20on%20this%20page.", "https://apiserpent.com/blog/serp-api-pricing-comparison#:~:text=This%20makes%20subscription%20models%20particularly%20wasteful.%20A,(100%2C000%2B%20searches)%20remain%20within%20typical%20academic%20budgets." ], "model": null, "provider_timestamp": null, "web_search": null, "search_queries": null, "served_by": null, "n_sources": 9, "grade": null, "brands": [ "DataForSEO", "Ahrefs", "Semrush", "SerpApi", "Search Console", "Bright Data", "Oxylabs" ] }, { "id": "P25", "route": "gemini", "latency_ms": 40405, "fetched_at": "2026-10-07T22:39:39Z", "answer": "Connecting Google Search Console (GSC) and keyword research tools directly to Claude Code (or the broader Claude ecosystem using the Model Context Protocol, or MCP) transforms Claude from a static writing assistant into an automated SEO command center.\n\nThe primary tools, connectors, and protocols used for this integration are organized below by approach:\n\n1. Model Context Protocol (MCP) Servers\n\nMCP is the open standard that allows Claude to securely interact with external tools and APIs without manual CSV exports.\n\nAhrefs MCP Server (@ahrefs/mcp):\n\nWhat it does: Gives Claude direct, structured access to backlink profiles, keyword volumes, difficulty scores, and site audits.\n\nBest for: Technical keyword research and competitive intelligence without leaving your coding/CLI workflow.\n\nGoogle Search Console MCP (Community & Managed):\n\nWhat it does: Pipes queries, impressions, click-through rates (CTR), and average positions straight into Claude. Open-source community options (such as mcp-server-gsc or specialized creator MCPs) or managed connectors bridge the GSC API directly to Claude via local configuration.\n\nBest for: Developers and technical SEOs using Claude Code who want native terminal-based access.\n\n2. No-Code Data Connectors & ETL Layers\n\nIf you prefer managed middleware that handles OAuth authentication, pagination, and data formatting automatically, several platforms offer native Claude connectors:\n\nSupermetrics for AI Chats:\n\nWhat it does: Provides a native connector link in the Claude directory to pipe GSC data (along with 170+ other marketing metrics) into your sessions.\n\nBest for: Cross-referencing search console data cleanly against Google Ads, GA4, or CRM platforms.\n\nPorter Metrics & Coupler.io:\n\nWhat they do: Act as middleware integration layers that connect GSC to Claude via custom connectors, offering pre-built analysis workflows for striking-distance keywords, cannibalization checks, and traffic drops.\n\n3. Purpose-Built SEO & GEO Platforms\n\nxSeek CLI:\n\nWhat it does: Specifically designed for AI visibility optimization and Generative Engine Optimization (GEO). It ships with built-in Claude Code skills (like /aeo-audit and /find-opportunities) to track how your brand appears across AI engines alongside traditional Google data.\n\nDataForSEO:\n\nWhat it does: Offers both an MCP server and a REST API supporting deep keyword research, SERP features, and AI Overview tracking on a pay-per-task model.\n\nChoosing the Right Approach\n\nIntegration Method\nSetup Effort\nBest For\nOpen-Source / Custom MCP (e.g., GSC / Ahrefs MCP)\nMedium (Node.js/Config)\nDevelopers using Claude Code who want direct API tool-calling.\nManaged Connectors (Supermetrics, Porter, Coupler)\nLow (No-code OAuth)\nMarketers and teams wanting reliable, zero-maintenance data syncing.\nSEO Suites (Ahrefs, xSeek)\nLow to Medium\nAdvanced keyword planning, backlink tracking, and AI visibility auditing.", "sources": [ "https://www.xseek.io/blogs/articles/best-seo-tools-integrate-claude-code", "https://portermetrics.com/en/tutorial/claude/chat-google-search-console/#:~:text=A%20Claude%20account%20%E2%80%94%20the%20free%20plan,Google%20Search%20Console%20work%20under%20the%20hood.", "https://www.xseek.io/blogs/articles/best-seo-tools-integrate-claude-code#:~:text=*%202.%20Ahrefs%20MCP%20Server%20%E2%80%94%20Backlinks,%E2%80%94%20crawl%20errors%2C%20broken%20links%2C%20redirect%20chains.", "https://www.xseek.io/blogs/articles/best-seo-tools-integrate-claude-code#:~:text=No%20more%20CSV%20exports%20%E2%80%94%20Claude%20queries,API.%20Keyword%20volumes%20and%20difficulty%20at%20scale.", "https://www.xseek.io/blogs/articles/best-seo-tools-integrate-claude-code#:~:text=Best%20for%3A%20Teams%20already%20on%20Ahrefs%20who,performance%20data.%20Index%20coverage%20and%20crawl%20stats.", "https://blog.coupler.io/how-to-connect-google-search-console-to-claude/#:~:text=Claude%20Google%20Search%20Console%20connector%20(MCP)%20There,dimensions%20like%20query%2C%20page%2C%20device%2C%20and%20country.", "https://supermetrics.com/connect/google-search-console-to-claude#:~:text=Zero%20code.%20*%20Open%20Supermetrics%20in%20Claude.,English.%20No%20queries%2C%20no%20exports%2C%20no%20waiting.", "https://supermetrics.com/connect/google-search-console-to-claude#:~:text=Absolutely.%20Supermetrics%20connects%20170%2B%20data%20sources%20to,ad%20spend%20with%20revenue%20or%20web%20analytics.", "https://blog.coupler.io/how-to-connect-google-search-console-to-claude/#:~:text=Coupler.io%20includes%20pre%2Dbuilt%20analysis%20skills%20that%20cover,Sheets%2C%20Looker%20Studio%2C%20BigQuery%2C%20or%20Power%20BI.", "https://www.xseek.io/blogs/articles/best-seo-tools-integrate-claude-code#:~:text=xSeek%20ships%20with%206%20built%2Din%20Claude%20Code,Monitor%20brand%20mentions%20across%20all%20AI%20engines.", "https://www.xseek.io/blogs/articles/best-seo-tools-integrate-claude-code#:~:text=*%204.%20DataForSEO%20%E2%80%94%20SERP%20Data%20%2B,tracking%20%E2%80%94%20which%20queries%20trigger%20AI%2Dgenerated%20answers." ], "model": null, "provider_timestamp": null, "web_search": null, "search_queries": null, "served_by": null, "n_sources": 11, "grade": null, "brands": [ "DataForSEO", "Ahrefs", "Search Console" ] }, { "id": "P26", "route": "gemini", "latency_ms": 52850, "fetched_at": "2026-10-07T22:41:10Z", "answer": "When looking to research competitors across Meta (Facebook & Instagram) and Google (Search, YouTube, and Display) ad libraries via APIs, you face a unique ecosystem constraint: Neither Meta nor Google offers a robust, developer-friendly public API explicitly built for competitor ad research or intelligence.\n\nMeta deprecated much of its public Ad Library API access for general competitive intelligence years ago (limiting it mostly to academic researchers and verified political/social issue transparency), and Google does not provide an official API for the Google Ads Transparency Center.\n\nHowever, depending on your technical requirements, budget, and scale, several options bridge this gap\u2014ranging from multi-platform intelligence APIs to enterprise scraping and creative analytics suites.\n\n1. Multi-Platform Intelligence APIs & Platforms\n\nIf you need programmatic access (via REST API or robust third-party endpoints) to pull data spanning both Meta and Google/YouTube, specialized ad intelligence software is usually your best path:\n\nAdLibrary.com API\n\nCoverage: Meta (FB/IG), Google, YouTube, LinkedIn, TikTok, Pinterest, Snapchat.\n\nBest For: SaaS, B2B, and multi-channel brands wanting a unified REST API key to pull text, media assets, and active/historical ad trends across both major ad ecosystems without managing separate scrapers.\n\nBigSpy / PowerAdSpy\n\nCoverage: Massive multi-network databases spanning Meta, Google/Display, TikTok, and more.\n\nBest For: Budget-conscious teams looking for out-of-the-box UI searches or basic data feeds to monitor high-level competitor ad volume, duration, and geographic targeting.\n\n2. Meta-Focused APIs & Creative Intelligence Suites\n\nIf your primary focus is deep-diving into Meta ad creatives, landing pages, and scaling patterns (with varying levels of programmatic or direct integrations):\n\nGetHookd\n\nCoverage: Meta, TikTok, YouTube Shorts.\n\nBest For: E-commerce and agency workflows. It provides developer-friendly capabilities (including API/MCP access on standard plans) paired with AI-driven ad script generation and brand spying tools.\n\nAdSpy\n\nCoverage: Exceptionally deep database for Facebook and Instagram.\n\nBest For: Granular text, URL, and landing page searching within Meta\u2019s ecosystem, though it lacks robust multi-platform Google coverage or a broad public developer API.\n\nAtria / Foreplay\n\nCoverage: Meta-heavy creative research.\n\nBest For: Creative strategists. While less focused on raw developer APIs and more on UI/swipe files and AI brief-generation, they are gold standards for analyzing how Meta competitors construct winning funnels.\n\n3. Google Ads Competitor Research (Search & YouTube)\n\nBecause Google does not offer an official Ads Transparency Center API, programmatic research on Google Search and YouTube ads typically relies on one of two approaches:\n\nSERP & Transparency Scraper APIs (e.g., Scrapeless, SerpApi, or custom proxy scrapers):\n\nDevelopers often use structured SERP scraping actors tailored to query the Google Ads Transparency Center or Google Search results programmatically. This allows you to extract competitor domain visibility, ad copy variations, and historical keyword footprint data at scale while bypassing CAPTCHAs and IP blocks.\n\nSEM/SEO Intelligence APIs (SEMrush API, Similarweb API, DataForSEO):\n\nIf your goal on Google is to see what Search keywords competitors are bidding on and what their text ad history looks like, traditional SEO/SEM data providers offer robust REST APIs that infer Google Ads strategies far more reliably than scraping the Transparency Center UI.\n\nSummary Recommendation\n\nIf you need a unified API covering both Meta and Google/YouTube: Look into multi-channel data providers like AdLibrary.com.\n\nIf you are building custom data pipelines for Google Search/Transparency data: Utilize enterprise SERP/scraping infrastructure providers (like SerpApi or specialized actors) to parse the public interfaces safely.\n\nIf you want turnkey creative analysis for Meta: Tools like GetHookd (which includes developer-accessible endpoints) or Atria will save you hundreds of hours compared to building custom Meta parsers.", "sources": [ "https://adlibrary.com/posts/best-meta-ad-library-alternatives-for-saas-2026#:~:text=Among%20meta%20ad%20library%20alternatives%20for%20SaaS%2C,manual%20cross%2Dplatform%20work%20and%20offer%20no%20export.", "https://adlibrary.com/posts/best-meta-ad-library-alternatives-for-saas-2026#:~:text=*%20YouTube%20%E2%80%94%20thought%20leadership%20and%20case,and%20lookalike%20audiences%20based%20on%20sign%2Dup%20behavior.", "https://adlibrary.com/posts/best-meta-ad-library-alternatives-for-saas-2026#:~:text=Among%20meta%20ad%20library%20alternatives%20for%20SaaS%2C,manual%20cross%2Dplatform%20work%20and%20offer%20no%20export.", "https://adlibrary.com/posts/best-meta-ad-library-alternatives-for-saas-2026#:~:text=Among%20meta%20ad%20library%20alternatives%20for%20SaaS%2C,manual%20cross%2Dplatform%20work%20and%20offer%20no%20export.", "https://www.gethookd.ai/learn/4-best-meta-ads-competitor-analysis-tools-in-2026/#:~:text=Atria%20connects%20to%20your%20Meta%20and%20TikTok,plus%20AI%20creative%20production%20on%20every%20plan.", "https://www.gethookd.ai/learn/4-best-meta-ads-competitor-analysis-tools-in-2026/#:~:text=GetHookd%20is%20a%20competitor%20ad%20research%20and,eCommerce%20operators%20running%20Meta%20and%20TikTok%20ads.", "https://www.gethookd.ai/learn/4-best-meta-ads-competitor-analysis-tools-in-2026/#:~:text=Saved%20ads%20remain%20in%20your%20library%20permanently%2C,the%20same%20competitor%20research%20an%20agency%20does.", "https://www.adstellar.ai/blog/meta-ad-library-scraping-tool#:~:text=Best%20for%3A%20Deep%20Facebook%20and%20Instagram%20ad,copy%2C%20advertiser%20names%2C%20or%20landing%20page%20URLs.", "https://www.gethookd.ai/learn/4-best-meta-ads-competitor-analysis-tools-in-2026/#:~:text=Stop%20Guessing.%20Start%20Scaling.%20GetHookd%20reveals%20which,generation%20and%20creative%20cloning.%20What%20You%20Get%3A", "https://adlibrary.com/posts/best-meta-ad-library-alternatives-for-saas-2026#:~:text=Free%20but%20siloed%20(use%20as%20supplements%2C%20not,Foreplay%20%E2%80%94%20Meta%2Dfocused%20creative%20inspiration%2C%20excellent%20UX.", "https://www.scrapeless.com/en/blog/google-ads-transparency#:~:text=Scrapeless%20MCP%20Server%20Is%20Officially%20Live!%20Build,Step%205%3A%20Search%20by%20Advertiser%20or%20Website.", "https://www.scrapeless.com/en/blog/google-ads-transparency#:~:text=You%20can%20select%20to%20view%20image%20ads%2C,Collect%20Public%20Ad%20Data%20With%20an%20API.", "https://adlibrary.com/posts/best-meta-ad-library-alternatives-for-saas-2026#:~:text=Among%20meta%20ad%20library%20alternatives%20for%20SaaS%2C,manual%20cross%2Dplatform%20work%20and%20offer%20no%20export." ], "model": null, "provider_timestamp": null, "web_search": null, "search_queries": null, "served_by": null, "n_sources": 13, "grade": null, "brands": [ "DataForSEO", "Semrush", "SerpApi", "Similarweb", "Ads Transparency", "Foreplay", "AdSpy", "BigSpy", "Swipe Files" ] }, { "id": "F1", "route": "perplexity-routed", "latency_ms": 4152, "fetched_at": "2026-10-07T22:34:10Z", "answer": "The **multimodal filter** in Google Search Console shows how your site performs in **web searches that use an image as part of the search**, rather than only typed text.[1]\n\nIt includes traffic from:\n\n- Google Lens\n- Circle to Search on Android\n- Image uploads to Google Search\n- Chrome\u2019s \u201cSearch this image\u201d feature[2]\n\nYou can find it in **Performance \u2192 Search results \u2192 Search type \u2192 Web \u2192 Multimodal**. It reports metrics such as impressions, clicks, click-through rate, and position, with breakdowns by page, country, device, and date.[3]\n\nA key limitation is that **query data is unavailable**, because these searches primarily use images instead of text.[1] It is also separate from the **Google Images** search type: multimodal describes *how the search began*, not necessarily where the result appeared.[6]", "sources": [ "https://www.relevantaudience.com/seo/search-console-multimodal-search-type-lens-circle-to-search/", "https://scalevise.com/resources/google-search-console-multimodal-filter/", "https://searchrepublic.co.nz/blog/google-search-console-multimodal-search-filter/", "https://www.thekeyword.co/news/search-console-multimodal-search-report", "https://www.searchenginejournal.com/google-search-console-multimodal-filter/590781/", "https://www.digitalapplied.com/blog/search-console-multimodal-filter-lens-visual-search", "https://www.affiversemedia.com/google-search-console-multimodal-reporting-affiliates/", "https://almcorp.com/news/search-console-adds-multimodal-filter-image-searches/", "https://www.seroundtable.com/google-search-console-multimodal-search-type-filter-42156.html", "https://www.seroundtable.com/recap-09-24-2026-42158.html" ], "model": "sonar", "provider_timestamp": "2026-10-07 22:34:10 +00:00", "web_search": true, "search_queries": null, "served_by": "dataforseo.x.ai-optimization-perplexity-llm-responses-live", "source_kind": "model_api", "cost_usd": 0.005801, "n_sources": 10, "grade": "correct", "brands": [ "Search Console" ] }, { "id": "F2", "route": "perplexity-routed", "latency_ms": 3444, "fetched_at": "2026-10-07T22:34:13Z", "answer": "Yes. Ahrefs changed its API limits in September 2026:\n\n- On **September 22**, Ahrefs announced that **API and MCP usage limits increased**, with updated allowances varying by plan.[1]\n- The reported new monthly API-unit allowances for Lite, Standard, and Advanced plans were **200,000, 800,000, and 2,000,000**, respectively\u2014double the previous limits.[10]\n- Ahrefs also added a default **1,000-result `limit`** to the \u201cSE Metrics by country\u201d endpoint and changed Site Explorer\u2019s default `traffic_mode` from `static` to `adaptive`.[1]\n- The standard API request-rate limit remained **60 requests per minute**, subject to dynamic throttling.[6]\n\nSo, if you mean monthly usage quotas, **yes\u2014those were increased in September**. The September changes also included endpoint behavior and output-format changes.", "sources": [ "https://docs.ahrefs.com/en/api/docs/changelog", "https://releasebot.io/updates/ahrefs/ahrefs-api", "https://busyless.space/seo-apis/ahrefs", "https://ahrefs.com/blog/new-features/", "https://withtool.dev/posts/ahrefs-api-alternatives", "https://docs.ahrefs.com/en/api", "https://rankable.ai/blog/ahrefs-api-pricing", "https://www.anygen.io/showcase/ahrefs-api/index.html", "https://www.linkedin.com/posts/glen-allsopp-63084025_big-news-ahrefs-just-massively-increased-activity-7454886990908403712-BJCh", "https://prtimes.jp/main/html/rd/p/000000069.000157671.html" ], "model": "sonar", "provider_timestamp": "2026-10-07 22:34:13 +00:00", "web_search": true, "search_queries": null, "served_by": "dataforseo.x.ai-optimization-perplexity-llm-responses-live", "source_kind": "model_api", "cost_usd": 0.00581, "n_sources": 10, "grade": "correct", "brands": [ "Ahrefs" ] }, { "id": "F3", "route": "perplexity-routed", "latency_ms": 3286, "fetched_at": "2026-10-07T22:34:16Z", "answer": "OpenSEO is an **open-source, MIT-licensed alternative to tools such as Semrush and Ahrefs**. It can be self-hosted and combines keyword research, competitor analysis, backlink analysis, rank tracking, technical site audits, and AI-search visibility in one interface.[1][9][10]\n\nIts primary data source is **DataForSEO**. OpenSEO does not build its own full SEO index; users provide a DataForSEO API key and pay DataForSEO on a pay-as-you-go basis for the data they request, including search volumes, SERPs, rankings, and backlink information.[1][6]", "sources": [ "https://openseo.so/open-source-seo", "https://note.com/ai_driven/n/n759af2612760?hl=en", "https://xcloud.host/what-is-openseo-features-and-how-it-works/", "https://github.com/every-app/open-seo/blob/main/web/content/blogs/best-open-source-seo-tools.md", "https://xcloud.host/what-is-openseo-features-and-how-it-works/index.md", "https://github.com/every-app/open-seo/blob/main/README.md", "https://xcloud.host/best-open-source-seo-tools/", "https://openseo.so/about", "https://github.com/every-app/open-seo", "https://openseo.so/" ], "model": "sonar", "provider_timestamp": "2026-10-07 22:34:16 +00:00", "web_search": true, "search_queries": null, "served_by": "dataforseo.x.ai-optimization-perplexity-llm-responses-live", "source_kind": "model_api", "cost_usd": 0.005759, "n_sources": 10, "grade": "correct", "brands": [ "DataForSEO", "Ahrefs", "Semrush" ] }, { "id": "F4", "route": "perplexity-routed", "latency_ms": 3484, "fetched_at": "2026-10-07T22:34:20Z", "answer": "Jev is TypeSafe AI\u2019s flagship **\u201cSystem One\u201d decision model**. Instead of generating conversational text like an LLM, it evaluates supplied text or JSON (\u201cstate\u201d) against typed questions and returns structured results that software can use directly.[1]\n\nIt supports three main decision types:\n\n- **Choice:** select an option, such as which team should handle a support ticket.\n- **Score:** rate something against a defined rubric, such as urgency or quality.\n- **Noul:** estimate the probability that a yes/no statement is true.[4]\n\nJev is intended for classification, routing, moderation, verification, and other fast automated decisions. Its outputs include probabilities or confidence information, but it does not write prose, generate code, or function as a chatbot.[2]", "sources": [ "https://docs.typesafe.ai/introduction", "https://docs.typesafe.ai/introduction/coding-agents", "https://en.wikipedia.org/wiki/Jev_(AI_model)", "https://developers.cloudflare.com/ai/models/typesafe/jev/", "https://www.digitalocean.com/resources/articles/what-is-jev", "https://www.browserbase.com/blog/what-is-jev", "https://www.business-standard.com/technology/artificial-intelligence/what-is-jev-inside-the-new-ai-model-built-to-make-software-decisions-126092200452_1.html", "https://zapier.com/blog/jev/", "https://www.firecrawl.dev/blog/what-is-jev", "https://www.sanity.io/glossary/jev-typesafe-ai-model" ], "model": "sonar", "provider_timestamp": "2026-10-07 22:34:20 +00:00", "web_search": true, "search_queries": null, "served_by": "dataforseo.x.ai-optimization-perplexity-llm-responses-live", "source_kind": "model_api", "cost_usd": 0.005778, "n_sources": 10, "grade": "correct", "brands": [] }, { "id": "P02", "route": "perplexity-routed", "latency_ms": 4315, "fetched_at": "2026-10-07T22:34:47Z", "answer": "The cheapest practical option is a **pay-as-you-go domain enricher**, rather than a sales-intelligence platform with a monthly minimum. Apify\u2019s Domain Enricher is listed at about **$5 per 1,000 domains** and returns WHOIS, DNS, email-provider, technology-stack, social-profile, and website metadata without a subscription.[1]\n\nFor very small lists:\n\n- **Free DIY:** Extract the domain from work emails, then manually check company details using the website, LinkedIn, or public registries. This costs nothing but requires review.[9][10]\n- **Free API tiers:** Abstract API offers 100 free requests, while Diffbot is reported to provide up to 10,000 monthly credits with a five-calls-per-minute limit.[3][5]\n- **Low-cost batch enrichment:** Use a pay-per-result service such as Apify when you need structured technical and company data; it is likely the best value for occasional or irregular jobs.[1]\n- **Avoid monthly tools:** Datablist is inexpensive per domain, but its reported pricing starts around $25/month, so it does not meet the no-subscription requirement.[2]\n\n**Recommendation:** Normalize and deduplicate your domains first, test 20\u201350 records, then run the full CSV through a pay-as-you-go Apify actor. For basic firmographics only, try a free API tier before paying. Note that inexpensive domain enrichment generally provides public website and technical data\u2014not reliably verified employee counts, revenue, or contacts.", "sources": [ "https://apify.com/silentflow/domain-enricher", "https://cufinder.io/blog/what-are-the-best-company-name-to-domain-finder/", "https://geekflare.com/guides/best-company-enrichment-apis/", "https://cufinder.io/blog/data-enrichment-pricing/", "https://coldiq.com/blog/best-company-enrichment-apis", "https://prospeo.io/s/company-name-to-domain-api", "https://cufinder.io/blog/api/best-company-name-to-domain-apis/", "https://tomba.io/blog/company-data-enrichment-from-domains", "https://cufinder.io/blog/how-to-get-website-url-from-company-name-in-bulk/", "https://cufinder.io/blog/how-to-find-company-information-for-your-email-list/" ], "model": "sonar", "provider_timestamp": "2026-10-07 22:34:47 +00:00", "web_search": true, "search_queries": null, "served_by": "dataforseo.x.ai-optimization-perplexity-llm-responses-live", "source_kind": "model_api", "cost_usd": 0.005936, "n_sources": 10, "grade": null, "brands": [ "Apify", "Abstract API" ] }, { "id": "P21", "route": "perplexity-routed", "latency_ms": 6408, "fetched_at": "2026-10-07T22:34:26Z", "answer": "For an **in-house growth agent**, the best setup is usually a combination of **first-party Google APIs** and one commercial provider for competitive intelligence:\n\n| API | Best for | Why it fits an agent |\n|---|---|---|\n| **Google Search Console API** | Your site\u2019s actual clicks, impressions, CTR, queries, pages, and positions | Authoritative first-party performance data; essential for opportunity detection and content prioritization |\n| **Google Analytics Data API** | Organic conversions, revenue, engagement, and funnel attribution | Connects rankings and traffic to business outcomes |\n| **DataForSEO API** | Broad SEO datasets: SERPs, keywords, backlinks, competitors, and audits | Strong choice for raw, programmatic data and large-scale workflows; often better suited to custom agents than dashboard-first tools[1] |\n| **SerpApi** | Targeted Google and other search-engine result retrieval | Easier SERP extraction when the agent mainly needs rankings, features, local results, or competitor visibility[1] |\n| **Ahrefs API** | Backlinks, referring domains, organic keywords, and competitor research | Good fit when link intelligence and mature competitive datasets are priorities |\n| **Semrush API** | Domain, keyword, competitor, and advertising intelligence | Useful for broad market research, though licensing and endpoint costs should be checked carefully |\n| **SE Ranking API** | An all-in-one SEO stack with rankings, backlinks, audits, and AI-search visibility | Attractive when the team wants API access plus a ready-made operational dashboard[2][3] |\n\n## Recommended stack\n\n**Best default for most growth teams:**\n\n1. **Google Search Console API** \u2014 identify declining pages, high-impression/low-CTR queries, and ranking opportunities.\n2. **Google Analytics Data API** \u2014 rank opportunities by conversions or revenue rather than traffic alone.\n3. **DataForSEO** \u2014 obtain competitor SERPs, keyword volumes, SERP features, and backlink data.\n4. **PageSpeed Insights API** \u2014 monitor Core Web Vitals and performance issues.\n5. **OpenAI or another LLM** \u2014 turn detected opportunities into briefs, experiments, and prioritized tasks.\n\nChoose **SerpApi instead of DataForSEO** if the agent primarily needs reliable, easy-to-consume SERP results rather than a broad SEO data warehouse. Choose **SE Ranking** if minimizing vendor integration work matters more than maximizing raw-data flexibility.[1][2]\n\n## What the agent should automate\n\n- Find queries with high impressions, low CTR, and positions near the top 10.\n- Detect pages losing clicks or rankings week over week.\n- Compare competitor URLs and SERP features for priority keywords.\n- Identify pages receiving traffic but producing few conversions.\n- Find keyword cannibalization and content gaps.\n- Monitor new or lost backlinks.\n- Create technical SEO tickets from crawl and performance data.\n- Measure whether title, content, internal-linking, and schema changes improve outcomes.\n\n## Selection criteria\n\nPrioritize APIs with:\n\n- Location, language, device, and search-engine controls.\n- Historical data and stable identifiers.\n- Batch endpoints and asynchronous jobs.\n- Clear usage-based pricing and rate limits.\n- Raw JSON responses rather than only dashboard exports.\n- Reliable handling of Google SERP features.\n- Terms that permit storage, derived metrics, and internal model use.\n\n**Bottom line:** start with **Search Console + Analytics + DataForSEO**. Add **PageSpeed Insights** for technical signals and **SerpApi** only if you need simpler or more specialized SERP collection. This gives an in-house agent both authoritative first-party data and competitive market intelligence without depending on a single SEO vendor.[1][2]", "sources": [ "https://www.agentseo.dev/blog/best-seo-api-for-ai-agents", "https://storylab.ai/best-seo-data-apis/", "https://analytify.io/best-seo-api/", "https://stackademic.com/blog/guide-on-top-10-best-seo-apis-in-2026-what-you-need-to-know", "https://www.onsaas.me/blog/best-seo-api", "https://www.sitepoint.com/top-10-seo-data-apis-technical-guide-to-programmatic-search/", "https://growwer.com/blog/the-best-seo-apis/", "https://nextgrowth.ai/best-seo-api/", "https://diyai.io/ai-tools/seo/seo-data-apis/", "https://www.marketing91.com/best-seo-data-api/" ], "model": "sonar", "provider_timestamp": "2026-10-07 22:34:26 +00:00", "web_search": true, "search_queries": null, "served_by": "dataforseo.x.ai-optimization-perplexity-llm-responses-live", "source_kind": "model_api", "cost_usd": 0.006385, "n_sources": 10, "grade": null, "brands": [ "DataForSEO", "Ahrefs", "Semrush", "SerpApi", "SE Ranking", "Search Console" ] }, { "id": "P22", "route": "perplexity-routed", "latency_ms": 5772, "fetched_at": "2026-10-07T22:34:32Z", "answer": "You can cover most keyword-volume and backlink needs with a **free-tool workflow plus occasional pay-as-you-go access**, rather than maintaining several SEO subscriptions.\n\n## Recommended low-cost setup\n\n- **Google Search Console:** Get the actual queries, impressions, clicks, CTR, and average positions for your own site. It is the best source for finding pages that already have ranking potential.[2]\n- **Google Keyword Planner:** Use it for keyword ideas, monthly volume estimates, CPC, and seasonality. Without active ad spend, volumes may appear as broad ranges rather than exact figures.[2][9]\n- **Google Trends:** Compare demand over time, identify seasonality, and distinguish growing topics from declining ones. Its numbers are relative indexes\u2014not monthly search volumes.[2]\n- **Free backlink checkers:** Ahrefs\u2019 free Backlink Checker can show a limited sample of backlinks and referring domains for a site or URL.[4]\n- **Bing Webmaster Tools:** Use its free keyword and site-performance data as a secondary source, especially for query discovery.\n- **Google search results:** Use autocomplete, \u201cPeople also ask,\u201d related searches, and competitor pages to expand your keyword list without paying for a database.\n- **Occasional paid access:** Subscribe to Ahrefs, Semrush, Moz, or a similar tool for one month when you need a large competitor-backlink export, then cancel and work from the exported data.\n\n## Practical workflow\n\n1. Export your existing queries from Search Console.\n2. Add keyword ideas from Keyword Planner, autocomplete, forums, Reddit, YouTube, and competitor pages.\n3. Use Trends to remove seasonal distortions and compare closely related terms.\n4. Group keywords by search intent rather than treating each variation as a separate article.\n5. Check the top-ranking pages manually for content quality, domain strength, and backlink patterns.\n6. Use a free backlink checker for initial research.\n7. For a deeper competitor audit, buy a short-term subscription or credits, export referring domains and relevant link opportunities, then cancel.\n8. Track your actual results in Search Console instead of relying solely on third-party traffic estimates.\n\n## When pay-as-you-go makes sense\n\nIf you regularly need competitor data but not every month, SEO APIs and credit-based services can provide keyword, SERP, and backlink data without a recurring subscription; some are billed per request or per row.[1] This is usually more economical for occasional audits, small sites, and one-off research than maintaining multiple monthly tools.\n\n**Best default stack:** Search Console + Keyword Planner + Trends + free backlink checkers, with one temporary or pay-as-you-go paid tool for quarterly competitor audits. This provides the essentials at little or no recurring cost, while paid databases mainly add larger indexes, more precise difficulty metrics, and competitor coverage.[5]", "sources": [ "https://busyless.space/seo-apis", "https://www.digidop.com/blog/seo-keyword-research-tools", "https://intergrowth.com/seo/diy-tools/", "https://buffer.com/resources/best-free-seo-tools/", "https://kolonell.com/en/blog/free-paid-seo-tools-comparison-2026", "https://rankyak.com/blog/affordable-seo-tools", "https://www.w3era.com/blog/seo/best-free-seo-tools-list/", "https://www.redmondpie.com/what-i-think-about-the-6-best-backlink-apis-my-honest-comparison/amp/", "https://www.stackmatix.com/blog/free-ai-keyword-research-tools", "https://backlinko.com/tools/keyword" ], "model": "sonar", "provider_timestamp": "2026-10-07 22:34:32 +00:00", "web_search": true, "search_queries": null, "served_by": "dataforseo.x.ai-optimization-perplexity-llm-responses-live", "source_kind": "model_api", "cost_usd": 0.006204, "n_sources": 10, "grade": null, "brands": [ "Ahrefs", "Semrush", "Moz", "Search Console" ] }, { "id": "P23", "route": "perplexity-routed", "latency_ms": 5415, "fetched_at": "2026-10-07T22:34:38Z", "answer": "The **cheapest practical approach** is usually a scheduled script using a pay-as-you-go SERP API\u2014especially **DataForSEO\u2019s queued API**, reported at about **$0.60 per 1,000 SERP checks**. For real-time checks, providers such as FlyByAPIs may be cheaper at higher volumes, but pricing and reliability should be verified before committing.[1][6]\n\n## Recommended setup\n\n1. Store your keywords, target domain, location, language, device, and desired SERP depth.\n2. Run a weekly cron job, GitHub Actions workflow, or serverless function.\n3. Submit keyword batches to a queued SERP endpoint.\n4. Parse the returned organic results to find your domain\u2019s position.\n5. Save results in SQLite, Postgres, or a spreadsheet.\n6. Compare each week\u2019s position with the previous result and send alerts only for meaningful changes.\n\nWeekly checks reduce usage to roughly one-seventh of daily checks; for example, 1,000 keywords checked weekly require about 4,300 requests per month rather than 30,000.[9]\n\n## Cost example\n\nFor **1,000 keywords**, one location, one device, and weekly checks:\n\n- Approximately 4,333 SERP requests per month\n- DataForSEO queued pricing: about **$2.60/month** at $0.60 per 1,000 checks\n- Actual cost may be higher because of minimum deposits, result depth, retries, extra locations, or multiple search engines.[6][10]\n\n## Best choice by situation\n\n| Need | Cheapest sensible option |\n|---|---|\n| Lowest raw batch cost | DataForSEO queued SERPs, around $0.60 per 1,000 checks[6] |\n| Real-time synchronous results | A low-cost real-time SERP API such as FlyByAPIs or similar[1][4] |\n| No engineering or database work | Keyword.com or a managed rank-tracking API, though the flat monthly fee can exceed raw API usage[7] |\n| Very small volume | Use a provider\u2019s free tier or trial, if available; some services advertise several thousand free requests or records per month[3][8] |\n\nAvoid scraping Google directly with proxies unless the volume is substantial and you are willing to maintain CAPTCHA handling, parsing, retries, and location accuracy. A managed SERP API is usually cheaper overall once maintenance is included.[2][6]\n\nFor most projects, the lowest-cost architecture is: **DataForSEO queued API + weekly GitHub Actions job + SQLite/Postgres + Search Console for your own site\u2019s actual clicks and impressions**. Reserve daily checks for a small set of high-value keywords; check the rest weekly or monthly.[1][6]", "sources": [ "https://flybyapis.com/blog/best-serp-tools/", "https://www.olostep.com/blog/best-rank-tracking-api", "https://www.octoparse.com/blog/best-serp-api", "https://flybyapis.com/blog/cheapest-serp-api/", "https://searlo.tech/blog/best-rank-tracker-api-2026", "https://geonode.com/blog/best-rank-tracker-apis", "https://keyword.com/api-pricing/", "https://apyhub.com/blog/top-5-apis-improve-seo-game", "https://hasdata.com/blog/rank-tracking-apis", "https://diyai.io/ai-tools/seo/seo-data-apis/" ], "model": "sonar", "provider_timestamp": "2026-10-07 22:34:38 +00:00", "web_search": true, "search_queries": null, "served_by": "dataforseo.x.ai-optimization-perplexity-llm-responses-live", "source_kind": "model_api", "cost_usd": 0.006199, "n_sources": 10, "grade": null, "brands": [ "DataForSEO", "Search Console" ] }, { "id": "P25", "route": "perplexity-routed", "latency_ms": 5334, "fetched_at": "2026-10-07T22:34:43Z", "answer": "The best setup is usually **Google Search Console via MCP + Ahrefs or Semrush for competitive keyword data**, with **DataForSEO** as a lower-cost, API-first alternative.\n\n| Tool | Best for | What Claude Code can access |\n|---|---|---|\n| **Google Search Console MCP** | First-party performance data | Queries, clicks, impressions, CTR, average position, indexing status, and crawl issues[1][3] |\n| **Ahrefs MCP** | Broad SEO research | Keyword ideas, volume, difficulty, backlinks, competitor gaps, and site-audit findings[3][7] |\n| **Semrush API/MCP** | Competitive and market analysis | Keyword databases, competitor rankings, SERP data, traffic estimates, and technical SEO data[4] |\n| **DataForSEO MCP** | Programmable, pay-as-you-go research | Keyword volume, SERP analysis, backlinks, and competitor intelligence[6] |\n| **Puppeteer MCP** | Technical validation | Rendered-page checks, metadata, structured data, internal links, and browser-based audits[7] |\n\n## Recommended combinations\n\n- **Best overall:** Search Console MCP + Ahrefs MCP + Puppeteer MCP.\n- **Best budget/API setup:** Search Console MCP + DataForSEO.\n- **Best for agencies:** Search Console MCP + Semrush or Ahrefs, depending on existing subscriptions.\n- **Best for content optimization:** Search Console MCP + Ahrefs. Use Search Console for queries already generating impressions, then Ahrefs for related terms and competitor gaps.\n\nSearch Console and third-party keyword tools serve different purposes: Search Console provides actual performance data for your properties, while Ahrefs, Semrush, and DataForSEO provide broader keyword and competitor estimates.\n\n## How to connect them\n\n1. Create a Google Cloud project and enable the Search Console API.\n2. Authenticate with OAuth or create a service account and grant it read access to the relevant Search Console property.[1][6]\n3. Install a Search Console MCP server and add it to Claude Code\u2019s MCP configuration.\n4. Add the Ahrefs, Semrush, or DataForSEO MCP/API integration separately.\n5. Store API keys in environment variables rather than directly in configuration files.\n6. Test with a prompt such as: \n `Find pages with high impressions, CTR below 2%, and average position between 4 and 20 over the last 28 days.`\n\nFor most teams, the strongest starting point is **Search Console MCP plus Ahrefs MCP**: the first identifies real opportunities on the site, while the second supplies keyword expansion, competitor context, backlinks, and content-gap analysis.[3][7]", "sources": [ "https://www.get-ryze.ai/blog/how-to-combine-claude-with-search-console-for-seo-wins", "https://www.linkedin.com/posts/nicholasdulait_10-claude-code-seo-workflows-everything-activity-7437760521782628352-Mv79", "https://www.xseek.io/blogs/articles/best-seo-tools-integrate-claude-code", "https://www.xseek.io/blogs/articles/connect-seo-tools-claude-code-xseek-ahrefs-semrush-mcp", "https://www.linkedin.com/posts/cyriaclefort_10-claude-code-seo-workflows-everything-activity-7440481593602793472-8CMu", "https://thestacc.com/blog/build-ai-seo-agent-claude-mcp/", "https://discoveredlabs.com/blog/claude-code-seo-playbook", "https://databar.ai/blog/article/automate-seo-claude-code", "https://www.linkedin.com/posts/anson-binoy_seo-claudeai-aitools-activity-7437779972003356673-_lVr", "https://www.indiehackers.com/post/google-search-console-keyword-research-202-ai-impressions-in-48-hours-85b605f4f3" ], "model": "sonar", "provider_timestamp": "2026-10-07 22:34:43 +00:00", "web_search": true, "search_queries": null, "served_by": "dataforseo.x.ai-optimization-perplexity-llm-responses-live", "source_kind": "model_api", "cost_usd": 0.006179, "n_sources": 10, "grade": null, "brands": [ "DataForSEO", "Ahrefs", "Semrush", "Search Console" ] }, { "id": "P26", "route": "perplexity-routed", "latency_ms": 5900, "fetched_at": "2026-10-07T22:34:53Z", "answer": "For most teams, the best setup is **Meta\u2019s Ad Library API plus a specialized Google Ads Transparency data provider**. Google does not offer a broadly equivalent official public API for unrestricted competitor-ad discovery, while Meta provides structured access to its public Ad Library dataset.[1][2]\n\n## Best options\n\n| API or provider | Best for | Strengths | Limitations |\n|---|---|---|---|\n| **Meta Ad Library API** | Official Meta research | Free, structured access to ads across Facebook, Instagram, Messenger, and Audience Network; supports advertiser, keyword, country, and status filtering.[1][3] | Requires a Meta developer app; rate limits and review requirements can complicate production-scale collection. |\n| **Google Ads Transparency Center data providers** | Google Search, Display, YouTube, and Shopping competitor research | Covers Google\u2019s broader advertising ecosystem and can provide searchable advertiser and creative data.[1] | Google\u2019s public transparency data is less standardized and accessible than Meta\u2019s; quality depends heavily on the provider. |\n| **AdLibrary API** | Unified, production-ready monitoring | Enriched creative metadata, run-duration signals, cross-platform workflows, and less operational friction than using Meta directly.[9] | Paid; verify coverage, freshness, historical depth, and licensing before depending on it. |\n| **Commercial ad-intelligence APIs** | Large-scale historical analysis | Often combine Meta and Google data with OCR, landing-page metadata, creative categorization, spend estimates, and alerts. | Usually expensive, and estimated metrics may not be authoritative. |\n\n## Recommended architecture\n\n- **Use Meta\u2019s official API** for Meta-only research when cost and provenance matter most.\n- **Use a commercial provider** when you need one normalized schema across Meta and Google, historical archives, creative downloads, or automated monitoring.\n- **Use Google\u2019s Ads Transparency Center directly for validation**, especially when checking whether a provider\u2019s Google coverage is complete.\n- Store normalized fields such as advertiser, platform, country, first-seen date, last-seen date, creative type, copy, destination URL, media asset, and transparency-library ID.\n- Add OCR and landing-page snapshots yourself if the API does not provide them.\n\n## Selection checklist\n\nBefore choosing a vendor, verify:\n\n- Google coverage: Search, YouTube, Display, Shopping, and political ads\n- Historical depth versus active-ads-only access\n- US and international geographic coverage\n- API rate limits, pagination, webhooks, and bulk export\n- Creative image/video availability and permitted reuse\n- Advertiser/entity resolution across Meta and Google\n- Data freshness and first-/last-seen accuracy\n- Terms permitting competitive intelligence and internal storage\n\n**Practical recommendation:** start with **Meta Ad Library API + a Google-focused commercial API**. Choose a unified provider only if reducing engineering effort and gaining historical/enrichment features outweighs the additional cost and dependency.", "sources": [ "https://adlibrary.com/posts/ads-spy-guide-2026", "https://makometrics.com/blog/meta-ad-library-guide", "https://adlibrary.com/posts/best-ad-spy-tools-2026", "https://www.get-ryze.ai/blog/competitor-ads-analysis", "https://www.tryatria.com/blog/how-to-find-your-competitors-ads", "https://www.adyntel.com/blog/best-ad-spy-tools/", "https://www.socialfetch.dev/guides/ad-library-api", "https://awesome-copilot.github.com/skill/competitor-ad-intelligence/", "https://adlibrary.com/posts/best-advertising-intelligence-tools", "https://adlibrary.com/posts/best-ad-intelligence-tools-2026" ], "model": "sonar", "provider_timestamp": "2026-10-07 22:34:53 +00:00", "web_search": true, "search_queries": null, "served_by": "dataforseo.x.ai-optimization-perplexity-llm-responses-live", "source_kind": "model_api", "cost_usd": 0.006218, "n_sources": 10, "grade": null, "brands": [ "Meta Ad Library", "Ads Transparency" ] } ] }