Magica MCP

official

Magica is your all-in-one AI platform, offering 2500+ cutting-edge tools under a single subscription.

What can you do with Magica MCP?

  • Generate media — Ask your assistant to create images, video, audio, or speech through Magica's model endpoint.
  • Browse models — Have your assistant find a model and inspect its exact input schema before running a generation.
  • Monitor generations — Check the status of asynchronous model runs until they complete, fail, or are canceled.
  • Reuse media — Search recent generations, uploads, and reusable media catalogs to reference prior assets.

Documentation

Overview

Connect AI assistants like Claude, Cursor, and Codex to Magica using the Model Context Protocol.

Magica MCP connecting AI assistants to media generation models Magica MCP connecting AI assistants to media generation models

Magica provides a remote Model Context Protocol (MCP) server that lets AI assistants discover models, generate media, check run status, and reuse uploaded or generated assets. It supports the Streamable HTTP transport for session-aware clients and plain HTTP for simpler ones.


Endpoint

https://inference.magica.com/v1/mcp

Clients that support custom headers can use the same Bearer API key as the REST API. OAuth-only clients such as Claude Desktop discover Magica's sign-in flow automatically.


What you can do

Create images, video, audio, speech, and other model outputs. Find a model and inspect its exact input schema before execution. Check asynchronous model runs until they complete, fail, or are canceled. Search recent generations, uploads, and reusable media catalogs.

Next steps

Connect your client in under a minute. See all 14 model-visible MCP tools.

This documentation is built and hosted on Mintlify, a developer documentation platform.