Exa MCP
UpExa's official MCP server — a search engine built on neural embeddings, not a classic keyword index. It returns results as clean, readable content instead of snippets, plus page crawling and code/docs search. The web_search_exa, web_fetch_exa and agent_run tools cover search, extraction and agentic research. Freemium: works with or without a key, at a lower limit without one.
Why this server
- Embedding-based search matches meaning, not just keyword overlap
- web_fetch_exa returns clean page text — no need to parse HTML yourself
- agent_run runs a multi-step research task in one call instead of a chain of requests
- 4.7k stars and an official hosted endpoint — a mature, actively developed project
Usage examples
Prompt: Find recent articles on MCP's Streamable HTTP architecture and pull their content
Result: The agent called web_search_exa on the topic, then web_fetch_exa on the top results, returning clean article text.
Prompt: Run a mini research task on service X's competitors and build a comparison table
Result: The agent called agent_run with the task description — the server ran a series of searches on its own and returned a structured report.
Health
1 reports in the last 30 days · Last report: 25 Jul 2026
Reports by day (30 days)
| date | works | partial | broken |
|---|---|---|---|
| 2026-07-25 | 1 | 0 | 0 |
Server tools · 2
Collected by auto-check from tools/list — the real list, not the description.
web_fetch_exaRead a webpage's full content as clean markdown. Use after web_search_exa when highlights are insufficient or to read any URL. Best for: Extracting full content from known URLs. Batch multiple URLs in one call. Returns: Clean text content and metadata from the page(s).
web_search_exaSearch the web for any topic and get clean, ready-to-use content. Best for: Finding current information, news, facts, people, companies, or answering questions about any topic. Returns: Clean text content from top search results. Query tips: describe the ideal page, not keywords. "blog post comparing React and Vue performance" not "React vs Vue". Use category:people / category:company to search through Linkedin profiles / companies respectively. If highlights are insufficient, follow up with web_fetch_exa on the best URLs.
Reviews
No reviews yet.