How to Give Claude and ChatGPT Live Funding Data With Fundable MCP
Connect Fundable's MCP server to Claude or ChatGPT and query live funding rounds, investors, and companies in plain language. Setup in three steps, Pro+.
Published August 31, 2026 by Jacob Klionsky
TL;DR: Connect Fundable's MCP server to Claude or ChatGPT and your assistant becomes a live venture data layer: ask for funding rounds, investors, and companies in plain language and get answers from the dataset, with no SQL and no export. Setup is three steps and an OAuth sign-in, on Pro+.
If you research startups or investors, your assistant can reason but it cannot tell you who raised this week, because it has no live funding data. Fundable MCP gives it that, a live venture data layer you can query in plain language and push into your own tools. Fundable is a real-time VC funding and startup data platform, and MCP is how its data reaches your assistant.
What is Fundable MCP
MCP, the Model Context Protocol, is an open standard that lets an AI assistant call an external data source directly. Fundable MCP is a hosted server that exposes Fundable's dataset through that standard, so an assistant can query funding rounds, investors, companies, and the people behind them in real time. Fundable is a real-time VC funding and startup data platform, and MCP is how its data reaches your assistant without an export step. The server is remote and hosted, so there is nothing to install or run locally.
Ask an assistant instead of writing a query
Sales teams use funding data to time outreach, investors use it to source deals, and founders use it to map who is backing companies like theirs. Most of those questions are one-off: which fintech companies raised a Series A this month, or who backed a given company's last round. Answering them through the API means writing a request, and through the dashboard means clicking filters. An assistant connected to Fundable MCP answers them in plain language, and it holds the thread when you refine the question. It fits research and ad-hoc lookups, not scheduled pipelines.
Ways to reach Fundable data
There are three ways to get Fundable data, and they fit different jobs.
- The REST API, for scheduled pipelines and enrichment. See how to track newly funded startups with an API.
- A Clay HTTP API column, for enriching a table you already work in. See how to add funding signals to Clay and Apollo.
- Fundable MCP, for conversational questions inside Claude or ChatGPT. That is this post.
Pick MCP when the work is exploratory and you would rather ask than build.
What you need
- A Fundable account on Pro+, the plan that includes MCP access.
- A client that supports remote MCP connectors: Claude on desktop, claude.ai, or Claude Code, or ChatGPT.
- Your Fundable login, for the OAuth sign-in.
Connect Fundable MCP in three steps
1. Add the Fundable connector. In your client's connector settings, add Fundable's MCP server: https://mcp.tryfundable.ai. It is a hosted, remote server, so there is nothing to install.
2. Authorize with your Fundable account. The server uses OAuth, so you sign in through your browser with your Fundable login. Access requires an active Pro+ subscription.
3. Ask your first question. Type a funding question in plain language. The assistant queries the dataset and answers in the chat, and you can keep refining from there.
Watch the full flow, from connecting the server to the first answer:
Once it is connected, you can ask things like:
- US companies that raised a seed round in the last 12 months, backed by a16z or Y Combinator, with a Stanford founder
- Who led a given company's latest round, and who else took part
- Series B SaaS companies in London that raised this quarter
- Which funds have backed at least three developer-tool startups this year
From there you can combine the results with your own notes and push qualified companies into your CRM, so the assistant works as a sourcing workspace, not just a search box.
Give your assistant a live view of the funding market
Connect Fundable MCP and your assistant can answer funding questions the moment you ask them, straight from the dataset.