AI media generation - image, video, audio and 3D - via the Model Context Protocol. connect once and ask your assistant to generate.

add as a connector

ChatGPT

1. open Settings, then Apps (or Connectors), then Connect an app
2. choose MCP server and paste the endpoint url above
3. sign in with your OTOY account when the OAuth window opens
4. back in chat, ask for something, e.g. generate an image of a lighthouse at dusk

Claude (claude.ai / desktop)

1. open Connectors from the composer or settings, choose Add connector
2. pick MCP server and paste the endpoint url above
3. approve the OTOY OAuth sign-in, then use the tools in any conversation

Claude Code (CLI)

claude mcp add otoy-studio --transport http https://mcp.otoy.studio/mcp

then run /mcp and approve the server. the first tool use will walk you through the OTOY sign-in.

other hosts

any MCP client that supports streamable HTTP and OAuth can use the same url - Cursor, VS Code, Goose, litellm, mcp-remote and others. point it at:

https://mcp.otoy.studio/mcp

what you can do