Best by task · MCP servers
Best MCP servers for AI agents
Model Context Protocol servers that give coding agents access to GitHub, browsers, docs and more. Each row shows license, maintainer and latest release from GitHub.
- 11tools ranked
9 Tested · 2 Reported
- agent builderswho it is for
- Oct 26, 2026last updated
Last updated . Rankings are editorial; paid placements never affect them.
Get the weekly brief: which AI tools to update, try or skip. Free.
Free. Unsubscribe anytime. No spam. We email a confirmation link first. Privacy
Figma MCP server
Figma's MCP server that gives agents design context.
Verdict: Figma's own MCP server that gives coding agents access to design files.
Firecrawl
Web data API: search, scrape and crawl for AI agents.
Verdict: Web data API: search, scrape and crawl for AI agents.
MCP reference servers
The MCP project's collection of reference servers.
Verdict: The MCP project's collection of reference servers.
Context7
Up-to-date, version-specific library docs served to coding agents.
Verdict: Up-to-date, version-specific library docs served to coding agents.
GitHub MCP Server
GitHub's official MCP server for repos, issues and pull requests.
Verdict: GitHub's official MCP server for repos, issues and pull requests.
MCP Inspector
Official MCP server explorer with a CLI mode.
Verdict: The quickest way to check what an MCP server exposes; run it on a current Node or npm may quietly install version 0.15.0.
Supabase MCP
Connect Supabase projects to AI assistants.
Verdict: Connect Supabase projects to AI assistants.
Sentry MCP
Sentry's agentic tooling and MCP server for errors and traces.
Verdict: Sentry's agentic tooling and MCP server for errors and traces.
Playwright MCP
Microsoft's MCP server that lets agents drive a browser with Playwright.
Verdict: Microsoft's MCP server that lets agents drive a browser with Playwright.
Notion MCP Server
Official Notion MCP server for reading and editing workspace content.
Verdict: Official Notion MCP server for reading and editing workspace content.
Side by side: what we recorded for the top 8
Every cell is read from the tool’s own pages on the date shown on its entry. “Not recorded” means we have not verified it, not that it does not exist.
| Tool | Pricing | Licence | Platforms | Evidence | Verified |
|---|---|---|---|---|---|
| Figma MCP server | Partly verified, checked Oct 25, 2026. Included with Figma plans | Proprietary | Figma desktop and remote (per guide) | Reported | |
| Firecrawl | Partly verified, checked Oct 25, 2026. Freemium | AGPL-3.0 (per GitHub) | API, SDKs, MCP, self-host | Reported | |
| MCP reference servers | Official price page read, checked Oct 25, 2026. Free and open source | MIT to Apache-2.0 transition (per LICENSE file) | Node.js and Python servers | Tested | |
| Context7 | Not verified, checked Oct 25, 2026. Not verified | MIT (per GitHub) | MCP server and SDK | Tested | |
| Chrome DevTools MCP | Official price page read, checked Oct 25, 2026. Free and open source | Apache-2.0 (per GitHub) | Node.js with Chrome | Tested | |
| GitHub MCP Server | Official price page read, checked Oct 25, 2026. Free and open source | MIT (per GitHub) | Local and remote server (per README) | Tested | |
| MCP Inspector | Partly verified, checked Oct 25, 2026. Open source | NOASSERTION (see the repository) | Node | Tested | |
| Supabase MCP | Official price page read, checked Oct 25, 2026. Free and open source | Apache-2.0 (per GitHub) | Node.js | Tested |
Who each pick suits
- Figma MCP server: Good for: Developers generating code from Figma designs. Skip it if: You do not use Figma.
- Firecrawl: Good for: Agents that need clean web content. Skip it if: You need a license that avoids AGPL obligations for self-hosting.
- MCP reference servers: Good for: Learning MCP server patterns; Starting points for custom servers. Skip it if: You need a polished vendor-supported integration.
- Context7: Good for: Agents that hallucinate outdated library APIs. Skip it if: Your stack is internal-only code.
- Chrome DevTools MCP: Good for: Debugging and performance work in a real Chrome. Skip it if: You want cross-browser test automation.
- GitHub MCP Server: Good for: Agents that need to read and act on GitHub repositories. Skip it if: You do not use GitHub.
- MCP Inspector: Good for: Anyone building or evaluating an MCP server. Skip it if: You only use servers through a client and never inspect them.
- Supabase MCP: Good for: Agents managing Supabase projects and databases. Skip it if: You use a different backend.
Compare the options
- Firecrawl vs Context7: which should you use?
- MCP reference servers vs GitHub MCP Server: which should you use?
- Playwright MCP vs Chrome DevTools MCP: which should you use?
- Supabase MCP vs Notion MCP Server: which should you use?
Looking for something else? See all best-for categories, or the quick-answer tables.
Setup guides
FAQ
What is the best AI tool for MCP servers?
Our current top pick is Figma MCP server: Figma's MCP server that gives agents design context. Rankings are editorial and paid placements never affect them.
How are the picks on this page chosen?
Ranks are editorial judgement made by Agentic Weekly when an entry is written or updated. They are not benchmark results, user ratings, popularity counts or paid positions. We rank only the 139 tools we track, not the whole market. "#1" means top of the tools we cover for that task, not best overall. The reason for each rank is the one-line blurb shown beside it. Typical reasons are how well the tool's own docs say it fits the task and how actively it is maintained according to its dated release notes. There is no numeric scoring formula, and we have not run quality head-to-head comparisons between ranked tools; our research studies measure install size, startup time and speed for some tools, and those numbers do not set ranks. Where a rank looks like a verdict on quality, it is our reading of official sources, not a measured result. The Tested label does not raise a rank. Tested means a small smoke test (install plus a short task, usually on a small local model), not a benchmark. Reported and Tested entries are ranked from the same kind of evidence. Payment cannot change a rank. Featured listings are labelled and are never used to order a list. Prices and plan details are shown only with a date and an official source. Where we could only find a third-party figure, or could not read the official page, the entry says so. Ranks can change when a release, price or doc changes. Each page shows when it was last verified.
How often is this list updated?
Every week, when new releases are added to the feed. Each entry shows its release date and when it was last verified.
How to cite this page
Agentic Weekly, "Best MCP servers for AI agents (updated weekly)", last updated 2026-10-26, https://theagenticweekly.com/best/mcp-servers
Keep the Reported or Tested label when you quote a verdict. Plain-text version: https://theagenticweekly.com/best/mcp-servers.md. More for AI assistants: /ai, /tools.json, /llms.txt.
Free · weekly · about six minutes
Get The Weekly Verdict, free
What shipped this week, what is worth switching to, and what to ignore. We run what we can and label what we only read.
- One email a week. Each release gets one call: update now, worth a look, or skip.
- Every claim shows its evidence. Reported means we read the sources; Tested means we ran it, with the exact scope and the failures left in.
- A free AI Tools Cheat Sheet (web page and printable PDF) lands in your inbox when you confirm.
Free. Unsubscribe anytime. No spam. We email a confirmation link first and add you only after you click it. We use your email only to send this newsletter and never sell it. Privacy