---
title: "Docs MCP server"
description: "Give Claude Code, Cursor, VS Code or any MCP client search and read access to this documentation through the docs MCP server at https://platform.pixai.art/docs-mcp. It serves the docs only; it does not generate images."
canonical: "https://platform.pixai.art/en/docs/docs-mcp"
last-updated: "2026-11-16"
---

# Docs MCP server

Give Claude Code, Cursor, VS Code or any MCP client search and read access to this documentation through the docs MCP server at https://platform.pixai.art/docs-mcp. It serves the docs only; it does not generate images.



This documentation is available to AI coding agents through a remote [Model Context Protocol](https://modelcontextprotocol.io) server. Once connected, your agent can look up endpoints, parameters and examples itself instead of guessing them while it writes code against the PixAI API.

The server serves the documentation only. It cannot generate images or videos or call the PixAI API on your behalf; the code your agent writes does that, with your [API key](/en/docs/quick-start/enroll-in-api).

* **URL:** `https://platform.pixai.art/docs-mcp`
* **Transport:** Streamable HTTP
* **Authentication:** none, and it never spends credits.

Connect your editor or agent [#connect-your-editor-or-agent]

Claude Code [#claude-code]

```bash
claude mcp add --transport http pixai-docs https://platform.pixai.art/docs-mcp
```

Cursor [#cursor]

Add the server to `~/.cursor/mcp.json`, or to `.cursor/mcp.json` in your project:

```json
{
  "mcpServers": {
    "pixai-docs": {
      "url": "https://platform.pixai.art/docs-mcp"
    }
  }
}
```

VS Code [#vs-code]

Add the server to `.vscode/mcp.json` in your project:

```json
{
  "servers": {
    "pixai-docs": {
      "type": "http",
      "url": "https://platform.pixai.art/docs-mcp"
    }
  }
}
```

Other clients [#other-clients]

Any client that supports remote MCP servers over Streamable HTTP can connect with the URL above. For a client that only supports local (stdio) servers, bridge it with [`mcp-remote`](https://www.npmjs.com/package/mcp-remote):

```json
{
  "mcpServers": {
    "pixai-docs": {
      "command": "npx",
      "args": ["-y", "mcp-remote", "https://platform.pixai.art/docs-mcp"]
    }
  }
}
```

Tools [#tools]

| Tool          | What it does                                                                                                                    |
| ------------- | ------------------------------------------------------------------------------------------------------------------------------- |
| `search_docs` | Full-text search over every docs page, in English, Japanese or Korean. Returns the matching pages and sections with their URLs. |
| `read_doc`    | Returns one page as Markdown, including the request and response schemas of API reference pages.                                |
| `list_docs`   | Returns the docs index: every page with its URL and a one-line description, and links to the OpenAPI specs.                     |

Try asking your agent: &#x2A;"Using the pixai-docs MCP server, write a script that generates an image with the PixAI API and downloads it."*

Docs for AI agents without MCP [#docs-for-ai-agents-without-mcp]

The same content is available over plain HTTP:

* [`/llms.txt`](https://platform.pixai.art/llms.txt) lists every page with a short description, following the [llms.txt](https://llmstxt.org) convention. [`/llms-full.txt`](https://platform.pixai.art/llms-full.txt) has the full text of every page in one file.
* Every docs page has a Markdown version: add `.md` to its URL, for example [`/en/docs/faq.md`](https://platform.pixai.art/en/docs/faq.md), or request the page with an `Accept: text/markdown` header.
* The OpenAPI 3.1 spec of API v2 is at [`/openapi.json`](https://platform.pixai.art/openapi.json), for client generators and agents that read specs directly.
* Each page has **Copy Markdown** and **Open** buttons under its title, to paste the page into a chat or open it in ChatGPT, Claude or Cursor.
