Model Context Protocol

source

public(spec schema and docs in one repo plus reference SDKs in separate public repos)

license

Apache-2.0(OSI, new code/spec)+MIT(OSI, legacy contributions)

docs

CC-BY-4.0(documentation other than the specifications

core features withheld

no — spec schema and docs ship in the one LF/AAIF-governed repo and every reference SDK is a public modelcontextprotocol repo - there is no vendor tier to withhold anything for

Model Context Protocol's specification, schema and documentation live in one public repository, and every reference SDK is a separate public repository under the same project. No vendor sells an edition of the protocol beside them. The license splits by contribution: new code and specification contributions are under Apache-2.0 and earlier contributions stay under MIT, both OSI-approved.

• https://api.github.com/repos/modelcontextprotocol/modelcontextprotocol recorded 2026-08-12

  Repo metadata for the spec repo. It holds the specification, the protocol schema (TypeScript, emitted as JSON Schema) and the documentation; the reference SDKs are separate repos in the same public org. Nothing is withheld from it and there is no second, differently-licensed tree.

• https://en.wikipedia.org/wiki/Model_Context_Protocol recorded 2026-06-04

  Anthropic origin; donated to Linux Foundation AAIF Dec 2025

• https://github.com/modelcontextprotocol/modelcontextprotocol/blob/main/LICENSE recorded 2026-06-04

  Apache-2.0 (new code/spec) + MIT (legacy) + CC-BY-4.0 (docs) license text

• https://raw.githubusercontent.com/modelcontextprotocol/modelcontextprotocol/main/LICENSE recorded 2026-09-28

  The three-way split: "All new code and specification contributions to the project are licensed under Apache-2.0. Documentation contributions (excluding specifications) are licensed under CC-BY-4.0." Pre-transition work stays MIT, and "No rights beyond those granted by the applicable original license are conveyed for such contributions."

• https://raw.githubusercontent.com/modelcontextprotocol/modelcontextprotocol/main/README.md recorded 2026-09-28

  The spec repository README lists what it holds - "MCP specification", "MCP protocol schema" and "Official MCP documentation" - with the schema "[defined in TypeScript](schema/2026-07-28/schema.ts)". Its License line still reads "This project is licensed under the [MIT License](LICENSE)." while the LICENSE file carries the transition terms. Searched the body for pricing, paid, enterprise, license key and hosted: none.

OpenAI uses Model Context Protocol across ChatGPT, its Agents SDK and the Responses API, Google DeepMind supports it in Gemini, and Microsoft's Semantic Kernel and Azure OpenAI integrate with it. The protocol has no package of its own, so adoption is measured by downloads of its two reference SDKs, `mcp` on PyPI and the TypeScript SDK's `@modelcontextprotocol/sdk` on npm, each also listed on the map as its own entry.

• https://api.npmjs.org/downloads/point/last-month/@modelcontextprotocol/sdk recorded 2026-09-28

  The npm downloads API returns "downloads":202740901 for "@modelcontextprotocol/sdk" over the window "start":"2026-08-29","end":"2026-09-27".

• https://en.wikipedia.org/wiki/Model_Context_Protocol recorded 2026-09-28

  The article states "Following its announcement, the protocol was adopted by major AI providers, including" OpenAI and Google DeepMind, says MCP "can be integrated with" Microsoft "Semantic Kernel" and Azure OpenAI, and records the December 2025 donation to the Agentic AI Foundation.

• https://pypistats.org/api/packages/mcp/recent recorded 2026-09-28

  "last_month":219043071,"last_week":53929472

• https://zuplo.com/blog/one-year-of-mcp recorded 2026-09-28

  Dated 2025-11-24: "OpenAI has adopted it across ChatGPT (sort of), their Agents SDK, and their Responses API." and "Google DeepMind announced MCP support for Gemini." It gives no SDK download figure.

MCP ships a specification, schema and documentation, and the repository publishes no benchmark, leaderboard or performance figure, so there is nothing to score for capability. Implementation breadth already shows up in the SDK downloads counted under adoption, and the Agent2Agent and Agent Client protocols are left unscored for the same reason.

• https://api.github.com/repos/modelcontextprotocol/modelcontextprotocol recorded 2026-08-14

  Repo metadata - description "Specification and documentation for the Model Context Protocol", topics docs / mcp / specification / standard, 8,950 stars, homepage modelcontextprotocol.io. What the project ships is a specification and its schema, and no benchmark result, leaderboard entry or performance figure is published for it.

• https://modelcontextprotocol.io/docs/getting-started/intro recorded 2026-09-29

  The docs introduction calls it "an open-source standard for connecting AI applications to external systems" and describes what it enables, with no benchmark, leaderboard or performance figure. Searched the body for benchmark and leaderboard: none.

• https://raw.githubusercontent.com/modelcontextprotocol/modelcontextprotocol/main/README.md recorded 2026-09-29

  The spec repository README states it contains the "MCP specification", the "MCP protocol schema" and "Official MCP documentation", with the schema "defined in TypeScript" first and made available as JSON Schema. Searched the body for benchmark, leaderboard, score and performance: none, so the repository publishes no benchmark result.