Postman FlowsBuild flowsAI

Use AI in Postman Flows

View as Markdown

Postman Flows supports two ways to bring AI into your workflows: AI request blocks, which let you interact with AI models directly inside a flow, and MCP server flows, which let you deploy a flow as a Model Context Protocol (MCP) server that AI agents can call as a tool.

AI request blocks

AI request blocks are flow building blocks that send prompts to an AI model of your choice — OpenAI, Anthropic, Google, or any other OpenAI-compatible provider — and pass the response to the next block in your flow. Use them to prototype AI-powered workflows, evaluate different models, or add AI-generated output to an existing flow.

To learn more, see Create AI request blocks and add them to your flows.

MCP server flows

MCP server flows are deployed flows that act as Model Context Protocol (MCP) servers. Each flow exposes a public URL that AI agents and MCP hosts — such as Claude Desktop, VS Code, or Cursor — can call as a tool. The flow processes the incoming request and returns the result generated by the connected AI model. You can build and run an MCP server without writing any code.

To learn more, see Create MCP servers with Postman Flows.