Postman’s MCP server puts the Postman API behind tools: workspaces, collections, environments, specifications and mock servers. It comes in several modes, so you choose between a small, reliable tool list and the full surface.
https://mcp.postman.com/minimal
Claude Sonnet 5
Chat with 60+ AI models on the same workflow — switch to a different model mid-conversation and re-run the same prompt, or use Compare mode to put several side-by-side and balance quality vs. cost.
OAuth (preferred) or a Postman API key on the US endpoints. The EU endpoints accept an API key only.
How models use it and what it is built for.
The server wraps the Postman API. In the default minimal mode it exposes the essential operations for workspaces, collections and environments. Full mode exposes all of the Postman API tools, more than 100 of them, covering specifications and API design, mock servers, monitors, documentation and comments. Code mode is aimed at generating API client code from a specification, and further modes cover learning resources and querying your API estate in natural language. The mode is chosen by URL, which is the lever worth using: a model choosing among a dozen tools picks correctly far more often than one facing a hundred, so start minimal and move to full only for the operations you are missing.
Typical tools an AI model can call. Exact names vary by version.
Taken from the official Postman documentation — see Postman MCP Server documentation for the full reference.
Examples
Minimal (US)
https://mcp.postman.com/minimalDefault. Essential tools only — start here.
Full (US)
https://mcp.postman.com/mcpAll Postman API tools, over 100.
Code (US)
https://mcp.postman.com/codeClient code generation from specs.
Full (EU)
https://mcp.eu.postman.com/mcpEU data residency. API key only — no OAuth.
Copy any of these into MCP Agent Studio after connecting.
List the collections in my Payments workspace and tell me which ones have no environment.
Create a staging environment that copies production but points base_url at staging.
Add a request to the Orders collection for GET /orders/{id} with an example response.
Spin up a mock server for the Orders collection and give me its URL.
This is not a single-model product: you get the same MCP connection with 60+ models (Claude, GPT, Gemini, DeepSeek, open-weight, and more), you can switch mid-conversation, and you can open Compare mode to run the same prompt against multiple models at once. The card above is a suggested starting point for this server — not the only choice.
Default pick for Postman
Claude Sonnet 5
In full mode the tool list is long and many tools overlap. Sonnet 5 is reliable at picking the specific collection or environment tool rather than a near neighbour.
Listing tools proves the server is reachable, not that a model can work with it. Evals go further: they read every tool on the server, write a test suite from its real schemas, and run it — code decides pass/fail on the responses (schema conformance, error codes, pagination, result caps) while a scoring model grades plain-English tasks driven through the tools.
Get a pass/fail report per tool with the evidence behind each verdict — and replay the same suite after every schema change. Destructive tools are excluded from the run.
Run evalsOpen MCP Agent Studio with the connection pre-filled. Add your token, pick any of 60+ models, and start chatting — no install required.
Open Agent StudioCommon questions about connecting, scoping and using it safely.
Postman’s own hosted MCP server, which exposes the Postman API as tools: workspaces, collections, environments, specs, mock servers, monitors and more.
Start with minimal, the default. Full mode exposes over 100 tools, which makes tool selection harder for every model. Switch only when you need operations minimal does not have.
On the US endpoints, OAuth is preferred and a Postman API key also works. The EU endpoints do not support OAuth, so use an API key there.
It manages Postman resources through the Postman API — collections, environments, mocks and monitors. To call an API directly from an agent, connect that API’s own MCP server, or test it here.
Postman says its remote servers gather anonymous telemetry, while the local server does not unless you opt in.
GitHub
Drive GitHub repos, PRs and issues with an AI agent.
Everything
The reference server that exercises every part of the MCP spec.
DeepWiki
Ask questions about any public GitHub repository and get answers grounded in its generated wiki.
Stripe
Query customers, subscriptions and payments with an AI agent.