
MCP Registry
Official discovery registry for Model Context Protocol servers, with a browsable UI and a public API.
In short
MCP Registry is the canonical index for Model Context Protocol servers, enabling agents and IDEs to discover external tools. It is best for developers building agent integrations or publishing their own MCP servers who need a reliable upstream source of metadata.
Pick MCP Registry if you're building agent tooling, an IDE integration, or publishing your own MCP server and need the upstream index.
Skip it if you want a curated, opinionated marketplace of vetted agent tools rather than a raw registry feed.
The Official MCP Registry is the canonical index of Model Context Protocol servers — the bridge that lets agents like Claude, Cursor, and other MCP-aware clients plug into external tools, data sources, and APIs. It runs as a hosted registry at registry.modelcontextprotocol.io with a browsable web UI (filter by latest versions, paginate through recently updated entries) and a documented REST API so other tools and IDEs can pull live server metadata.
It's primarily for developers building or integrating MCP servers: maintainers publish servers and versions here, and client builders (agent IDEs, orchestration frameworks, marketplaces) consume the registry to surface installable connectors. The project is open source on GitHub, free to use, and supports multiple environments (production, staging, local) so teams can run their own mirror or test against a local instance before publishing.
Because this is infrastructure rather than a polished end-user product, expect rough edges: it's a registry, not a curated marketplace, so quality and abandonment vary server-by-server. But if you're working anywhere in the MCP ecosystem, this is the upstream source of truth.
This is plumbing, not product — but it's important plumbing. As MCP becomes the de facto interop standard for agent tools, the official registry is where every serious client will eventually pull from. Worth knowing about even if you never visit the page directly.
— The AI Tool Bible editorial team
Pros
- ✅ Official, canonical source for MCP server metadata
- ✅ Open source with a public REST API for programmatic access
- ✅ Supports multiple environments (production, staging, local)
- ✅ Free to use and free to publish servers
Cons
- ⚠️ Bare-bones UI focused on discovery, not curation or reviews
- ⚠️ Quality and maintenance varies widely across listed servers
- ⚠️ Aimed at developers; not useful for non-technical end users
Use cases
Frequently asked
- Who is the MCP Registry designed for?
- It is primarily for developers building or integrating MCP servers, including maintainers who publish servers and client builders who consume registry data for IDEs or frameworks.
- Is the MCP Registry free to use?
- Yes, the project is free, open source, and allows users to publish servers at no cost.
- Does the registry offer a programmatic interface?
- Yes, it provides a documented REST API that allows other tools and IDEs to pull live server metadata.
- Can teams run their own instance of the registry?
- Yes, it supports multiple environments including production, staging, and local, allowing teams to run their own mirror or test locally.
- Is the MCP Registry a curated marketplace?
- No, it is a raw registry rather than a curated marketplace, meaning the quality and maintenance of listed servers vary.
Explore related
Compare with similar tools
All in Agents →
LangGraph
FeaturedStateful, graph-based agent orchestration from LangChain.

CrewAI
FeaturedPython framework for multi-agent orchestration.

Ernie Bot
Baidu's Mandarin-first ChatGPT rival, powered by the ERNIE model family

Moveworks
The enterprise AI assistant that searches, answers, and takes action across your business systems

AWS Bedrock
Build and scale generative AI applications with foundation models

Claude Agent SDK
Anthropic's official SDK for building autonomous Claude agents.