What it does
This MCP server lets an MCP-compatible client generate AI videos from a text prompt or from a reference image used as the starting frame. It supports five Wan models, three resolution options (480P for drafts, 720P as the default, and 1080P for higher quality), audio generation, and character transfer from reference videos with the wan2.6-r2v model. Its tools cover submitting text and image generation, querying one or several task statuses and results, and listing the available models, resolutions and API actions.
Use it when you want to create a short AI video from a written description or a still image directly inside an AI assistant chat or editor, without switching to a separate video tool. It also fits workflows that submit several generation jobs and need to check their progress and retrieve finished results.
There are three ways to connect. Hosted OAuth uses the remote endpoint and a browser sign-in to AceDataCloud, so no token is pasted into client configuration. A hosted API token is sent as a Bearer value in the Authorization header and is created from an AceDataCloud application, where expiration, usage limits and allowed APIs can be set. A local stdio route runs the mcp-wan package as a local process that receives the token through an environment variable and still calls the AceDataCloud API. Setup examples are given for Claude, Claude Code, Cursor, VS Code and Codex.
Limitations and requirements
An AceDataCloud account is required for browser sign-in and for creating API keys. A generation key must have access to the required service and model and needs a balance that covers the request; task lookup alone does not permit submitting a generation. The hosted service can be metered, so current pricing should be reviewed before a real operation. Do not configure an OAuth login and a fixed Authorization header for the same server. Dynamic Client Registration registers the client application and is not an API key. Tokens must be kept out of prompts, shared workflows, screenshots and Git commits.
Safety
Our team has not reviewed this resource. Only automated rules checked it; resources with high-risk signals are not published. These are signals, not a guarantee that installing it is safe. Try it in a sandbox project first.
- Runs code or changes an assistant's behaviour
Sources
- Official MCP Registry first seen 13 Nov 2026 · last seen 15 Nov 2026 · its page on the source
io.github.AceDataCloud/mcp-wan
MCP server for Wan AI video generation
v2026.11.15.0As shown on registry.modelcontextprotocol.io when we last read it