Connect AI assistants to your Finseo data through the Model Context Protocol. Ask Claude, ChatGPT, or Cursor how your visibility in ChatGPT, Perplexity, Gemini, and Google AI Overviews changed, which competitors are ahead, and what to fix next.
The Finseo MCP server exposes your tracking data as tools an AI assistant can call. Once connected, Claude, ChatGPT, or Cursor can run list_projects, get_visibility_metrics, get_visibility_timeseries, list_prompts, get_prompt_details, get_answer_content, get_competitor_ranking, get_competitor_gap_analysis, get_competitor_h2h, get_sentiment_overview, get_top_sources, get_tags, get_query_fanouts, list_tasks, and get_task_details. Every tool is read-only and returns the same values as the Finseo dashboard, including visibility rate, mention rate, citation rate, ordinal position, mention depth, Share of Voice, and sentiment. Analytics tools accept optional model filters (chatgpt, perplexity, ai_overview, copilot, gemini, grok) and tag filters, default to a 30-day window, and support absolute startDate and endDate ranges.
Ask Claude how your brand visibility changed in the last 30 days versus the previous period. It calls get_visibility_metrics and get_visibility_timeseries and explains the trend, including how much comes from being named versus being cited.
Ask ChatGPT to rank your competitors and run a gap analysis against the strongest one. The assistant chains get_competitor_ranking, get_competitor_gap_analysis, and get_competitor_h2h to show where competitors appear when you do not, and who gets named first.
Drill from list_prompts to get_prompt_details to get_answer_content and read what ChatGPT or Perplexity actually answered on a given day, including every brand named and every source cited, without opening the dashboard.
Ask for your most important open optimization tasks. list_tasks returns the Finseo Tasks backlog sorted by priority score, and get_task_details delivers the step-by-step plan and the evidence that triggered it.
Developers working on a site can ask Cursor which sources AI models cite for tracked prompts or which fan-out queries appear most, then implement the fix in the same session using an API key with read scope.
The Model Context Protocol has become the standard way to give AI assistants access to external data, and Claude, ChatGPT, and Cursor all support remote MCP servers. Teams already ask these assistants to summarize, compare, and plan; without a connector they have to paste screenshots or CSV exports of their AI visibility data. The Finseo MCP server removes that step. Answer Engine Optimization (AEO) and Generative Engine Optimization (GEO) questions get answered from live tracking data with explicit metric definitions, so the assistant never confuses mention depth with a ranking position or visibility with mention rate.
How using Finseo MCP on its own compares to pairing it with Finseo for AI visibility and attribution.
| Feature | Finseo MCP alone | Finseo MCP + Finseo |
|---|---|---|
| Getting data into an assistant | Screenshots and pasted CSV exports | 15 native tools the assistant calls itself |
| Authentication | Copying keys into JSON config | OAuth sign-in for Claude and ChatGPT |
| Drill-down | Overview numbers only | From KPI to the exact AI answer and its sources |
| Metric accuracy | Assistant guesses what a metric means | Labeled fields identical to the dashboard |
| Safety | Risk of accidental changes | Read-only tools, project restrictions enforced |
Getting data into an assistant
Screenshots and pasted CSV exports
15 native tools the assistant calls itself
Authentication
Copying keys into JSON config
OAuth sign-in for Claude and ChatGPT
Drill-down
Overview numbers only
From KPI to the exact AI answer and its sources
Metric accuracy
Assistant guesses what a metric means
Labeled fields identical to the dashboard
Safety
Risk of accidental changes
Read-only tools, project restrictions enforced
In Claude, open Settings, Connectors, Add custom connector. Enter Finseo as the name and https://api.finseo.ai/v1/mcp as the Remote MCP Server URL, leave Client ID and Secret empty, click Add, then Connect, sign in with your Finseo account, and click Allow access.
Turn on Developer mode under Settings, Security and login. Open Settings, Plugins, click the plus button, enter Finseo and the MCP server URL https://api.finseo.ai/v1/mcp, click Create, complete the Finseo sign-in, and wait for the tool scan to finish.
Create an API key with read scope under Settings, API. In Cursor, add an MCP server with the URL and an Authorization: Bearer header. In Claude Code, run claude mcp add --transport http finseo https://api.finseo.ai/v1/mcp with or without the header.
Try "How visible is my brand in AI search this month, and how did it change versus the previous period?" or "Rank my competitors by AI visibility and run a gap analysis against the strongest one".
Follow up with prompt-level questions. The assistant chains list_prompts, get_prompt_details, and get_answer_content to show the exact answer text, brands named, and sources cited for any tracked prompt and date.
Finseo offers a native MCP server at https://api.finseo.ai/v1/mcp that connects Claude, ChatGPT, and Cursor to your AI visibility data through the Model Context Protocol. Fifteen read-only tools return visibility, competitor, sentiment, and citation data from ChatGPT, Perplexity, Gemini, Google AI Overviews, Microsoft Copilot, and Grok tracking, so Answer Engine Optimization (AEO) questions get answered from live numbers.
Add an MCP server as a custom connector in Claude, sign in, and ask in plain language, for example how your brand visibility in ChatGPT and Perplexity changed in the last 30 days. Claude calls get_visibility_metrics and get_visibility_timeseries on the Finseo MCP server and explains the trend, including how much comes from being named versus cited.
Yes. Finseo is a Generative Engine Optimization (GEO) platform whose remote MCP server works with Claude, ChatGPT in developer mode, Cursor, and Claude Code. The assistant and Finseo do different jobs: Finseo tracks how ChatGPT, Perplexity, and Gemini mention and cite your brand, and the assistant reads that data through MCP tools to summarize, compare, and plan.
In Claude, open Settings, Connectors, Add custom connector, enter Finseo and https://api.finseo.ai/v1/mcp, leave Client ID and Secret empty, click Connect, and allow access after signing in. In ChatGPT, enable Developer mode, then Settings, Plugins, add the same URL, and complete the Finseo sign-in. Both use OAuth 2.1; Cursor uses an API key with read scope instead.
Assistants read the same values as the Finseo dashboard: visibility rate, mention rate, citation rate, ordinal position, mention depth, Share of Voice, sentiment, prompts, answers, sources, fan-outs, and tasks. Nothing is ever modified because every MCP tool is read-only. Workspace project restrictions are enforced automatically, and you can revoke access by disconnecting the connector.
The Finseo MCP server has no separate fee, but you need a Finseo account on a paid plan with at least one tracking project. On the client side, Claude custom connectors work on Free, Pro, Max, Team, and Enterprise plans, ChatGPT requires Pro, Plus, Business, or Enterprise/Edu with developer mode, and Cursor only needs an API key.
Yes. Developers add the Finseo MCP server in Cursor with the URL https://api.finseo.ai/v1/mcp and an Authorization: Bearer header carrying an API key with read scope. They can then ask which sources AI models cite for tracked prompts or which fan-out queries appear most, and implement the fix in the same session. Claude Code works the same way.
Without MCP, Claude or ChatGPT only knows what you paste: screenshots or CSV exports, and it guesses what each metric means. The Finseo MCP server gives the assistant live tracking data across ChatGPT, Perplexity, and Gemini with explicit metric definitions, so it never confuses mention depth with a ranking position. Generative Engine Optimization (GEO) needs measured data, not anecdotes.
Every Finseo MCP analytics tool defaults to a relative 30-day window ending yesterday via timeframeDays. For an exact period, pass startDate and endDate together in YYYY-MM-DD format; get_visibility_metrics then compares that range against the same number of days before it. get_visibility_timeseries supports a maximum of 90 days. Model filters accept chatgpt, perplexity, ai_overview, copilot, gemini, and grok.
Pull every AI visibility KPI from Finseo into your own tools, data warehouse, or BI stack. Daily timeseries, prompt-level results with citations, competitor rankings, top cited sources, query fan-outs, tags, tasks, attribution, and bulk exports, all from ChatGPT, Perplexity, Claude, Gemini, and Google AI Overviews tracking.
View integration
Import your Finseo AI search visibility data straight into Google Looker Studio with the native community connector. Build custom dashboards and white-label client reports from ChatGPT, Perplexity, Gemini, Grok, and Google AI Overviews data.
View integration
Add Finseo visibility tasks as pages to any Notion database, such as your content calendar or sprint board, with the full recommendation as page content, so your team plans the fixes behind ChatGPT, Perplexity, and Gemini citations.
View integrationJoin hundreds of brands using Finseo to track and improve their visibility in ChatGPT, Perplexity, Claude, Gemini, and Google AI Overviews.
