Enables AI agents to generate images and videos, poll task status, and check pricing using the Wan model line on RunAPI.
Also known as runapi-ai/wan-mcp, @runapi.ai/wan-mcp
- Stars
- 0
- Tools
- 8
- Downloads, 30 days
- —
- Latest version
- 0.2.2
- Last commit
- Oct 22, 2026
- License
- Apache-2.0
What is Wan MCP?
Wan enables AI agents to generate images and videos, poll task status, and check pricing using the Wan model line on RunAPI. It exposes 8 tools, such as animate, edit_video, image_to_video.
Summary generated from the project's README
About Wan MCP server
Generating images and videos through the Wan model line is what this MCP server enables for AI agents. It provides direct access to 18 model variants across 6 specific endpoints, allowing assistants to perform creative tasks without needing to navigate a full model catalog. Users can integrate this into environments like Claude Code, Cursor, Windsurf, VS Code, and Roo Code using standard stdio commands.
Developers can use tools such as text_to_image, text_to_video, animate, and image_to_video to create media. The server also includes get_task for polling asynchronous results and check_pricing to verify costs before starting a job. Authentication is handled via a RUNAPI_API_KEY environment variable for headless environments or through a login tool that saves credentials to a local config file. This setup ensures that agents can manage the entire lifecycle of a media generation task from initial pricing checks to retrieving final output URLs.
What you can do with it
- 1Animate existing imagesA developer can instruct an assistant to transform a static image into a moving video.
- 2Monitor task progressAn agent can check the status of a long-running video generation job to see when it is finished.
- 3Verify model costsAn assistant can look up how much a specific Wan model will cost before executing a generation command.
Written from the project's README. Every use case is checked against a quote from it.
How to install Wan MCP
Formats verified 10 Nov 2026Add to CursorAdd to GooseInstall Wan MCP in Claude Code
$ claude mcp add --env 'RUNAPI_API_KEY=<YOUR_RUNAPI_API_KEY>' wan-mcp -- npx -y @runapi.ai/wan-mcpInstall Wan MCP in Cursor
~/.cursor/mcp.json
{
"mcpServers": {
"wan-mcp": {
"command": "npx",
"args": [
"-y",
"@runapi.ai/wan-mcp"
],
"env": {
"RUNAPI_API_KEY": "<YOUR_RUNAPI_API_KEY>"
}
}
}
}Install Wan MCP in VS Code
.vscode/mcp.json
{
"servers": {
"wan-mcp": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"@runapi.ai/wan-mcp"
],
"env": {
"RUNAPI_API_KEY": "<YOUR_RUNAPI_API_KEY>"
}
}
}
}Install Wan MCP in VS Code CLI
$ code --add-mcp '{"name":"wan-mcp","type":"stdio","command":"npx","args":["-y","@runapi.ai/wan-mcp"],"env":{"RUNAPI_API_KEY":"<YOUR_RUNAPI_API_KEY>"}}'Install Wan MCP in Claude Desktop
claude_desktop_config.json
{
"mcpServers": {
"wan-mcp": {
"command": "npx",
"args": [
"-y",
"@runapi.ai/wan-mcp"
],
"env": {
"RUNAPI_API_KEY": "<YOUR_RUNAPI_API_KEY>"
}
}
}
}Install Wan MCP in Codex
$ codex mcp add wan-mcp --env 'RUNAPI_API_KEY=<YOUR_RUNAPI_API_KEY>' -- npx -y @runapi.ai/wan-mcpInstall Wan MCP in Codex config.toml
~/.codex/config.toml
[mcp_servers.wan-mcp] command = "npx" args = ["-y", "@runapi.ai/wan-mcp"] [mcp_servers.wan-mcp.env] RUNAPI_API_KEY = "<YOUR_RUNAPI_API_KEY>"
Install Wan MCP in Gemini CLI
$ gemini mcp add -s user -e 'RUNAPI_API_KEY=<YOUR_RUNAPI_API_KEY>' wan-mcp npx -- -y @runapi.ai/wan-mcpInstall Wan MCP in Gemini settings.json
~/.gemini/settings.json
{
"mcpServers": {
"wan-mcp": {
"command": "npx",
"args": [
"-y",
"@runapi.ai/wan-mcp"
],
"env": {
"RUNAPI_API_KEY": "<YOUR_RUNAPI_API_KEY>"
}
}
}
}Wan MCP tools 8
Detected from the READMEanimateedit_videoimage_to_videospeech_to_videotext_to_imagetext_to_videoget_taskcheck_pricing
| animate | — | |
| edit_video | — | |
| image_to_video | — | |
| speech_to_video | — | |
| text_to_image | — | |
| text_to_video | — | |
| get_task | — | |
| check_pricing | — |
Configuration
RUNAPI_API_KEYSecretOptional RunAPI API key for headless hosts. Interactive MCP hosts can use the login tool, which writes shared RunAPI config.
| Name | Description | |
|---|---|---|
| RUNAPI_API_KEY | Optional RunAPI API key for headless hosts. Interactive MCP hosts can use the login tool, which writes shared RunAPI config. | Secret |
Automated checks
Ran Nov 13, 2026Full report →Passed our automated checks. That is not a security audit: read the code and give the server the narrowest access it needs. How checks work
- Source repository linked and reachablePass
- Open source license (OSI)Pass
- Commit or release in the last yearPass
- Package name not close to a popular packagePass
- Remote endpoint uses HTTPSNot checked
- Destructive tools are labelledPass
- No hidden instructions to the modelPass
- No critical advisory in the package or its dependenciesPass
README
View on GitHubWhy This Package?
@runapi.ai/wan-mcp is a focused Model Context Protocol server for the Wan model line on RunAPI.
It gives MCP-compatible assistants direct access to 6 endpoints and 18 model variants without loading the full RunAPI catalog.
Use this per-model server when an agent should stay scoped to Wan. Use @runapi.ai/mcp when one assistant should discover every RunAPI model line.
Install
Add it to Claude Code:
claude mcp add wan -s user -- npx -y @runapi.ai/wan-mcp
Use project scope when the server should be shared with a repository:
claude mcp add wan -s project -- npx -y @runapi.ai/wan-mcp
Codex, Cursor, Windsurf, VS Code, Roo Code, and other MCP hosts can use the same stdio command:
{
"mcpServers": {
"wan": {
"command": "npx",
"args": ["-y", "@runapi.ai/wan-mcp"]
}
}
}
check_pricing works before sign-in. For task creation and status polling, ask your assistant to call the login tool. It opens a browser login and saves credentials to ~/.config/runapi/config.json, the same file used by runapi login.
Headless and CI hosts can still set RUNAPI_API_KEY before starting the MCP host.
Ready-made examples are in examples/ for Claude, Cursor, Windsurf, VS Code, and Roo Code.
Tools
| Tool | Auth | Purpose |
|---|---|---|
animate |
Yes | Create a Wan animate task and optionally wait for a terminal status. Returns the task id, status, and output URLs. |
edit_video |
Yes | Create a Wan edit video task and optionally wait for a terminal status. Returns the task id, status, and output URLs. |
image_to_video |
Yes | Create a Wan image to video task and optionally wait for a terminal status. Returns the task id, status, and output URLs. |
speech_to_video |
Yes | Create a Wan speech to video task and optionally wait for a terminal status. Returns the task id, status, and output URLs. |
text_to_image |
Yes | Create a Wan text to image task and optionally wait for a terminal status. Returns the task id, status, and output URLs. |
text_to_video |
Yes | Create a Wan text to video task and optionally wait for a terminal status. Returns the task id, status, and output URLs. |
get_task |
Yes | Fetch the current status and latest payload for an existing task. |
check_pricing |
No | Look up current pricing for a Wan model and endpoint. |
Models
Wan covers 18 model variants across 6 endpoints. Each tool accepts the models listed for it:
| Tool | Models |
|---|---|
animate |
wan-2.2-animate-move, wan-2.2-animate-replace |
edit_video |
wan-2.6-edit-video, wan-2.6-flash-edit-video, wan-2.7-edit-video |
image_to_video |
wan-2.2-a14b-image-to-video-turbo, wan-2.5-image-to-video, wan-2.6-flash-image-to-video, wan-2.6-image-to-video, wan-2.7-image-to-video |
speech_to_video |
wan-2.2-a14b-speech-to-video-turbo |
text_to_image |
wan-2.7-image, wan-2.7-image-pro |
text_to_video |
wan-2.2-a14b-text-to-video-turbo, wan-2.5-text-to-video, wan-2.6-text-to-video, wan-2.7-r2v, wan-2.7-text-to-video |
Model availability can change between releases. Use check_pricing or the Wan model page for the current catalog view.
Agent Prompts
Ask your assistant in natural language; it can inspect pricing, create the task, and return the task id plus output URLs.
Create a task
Run a Wan animate task with RunAPI.
The assistant can call check_pricing, then animate, and return the task id, status, and output URLs.
Submit without waiting
Create the task but don't wait for it to finish.
The assistant calls the create tool with wait: false and returns the task id. Check on it later with get_task.
Check pricing before creating
Check current Wan pricing, then create the task if it matches my request.
The assistant calls check_pricing and can link to the Wan model page for the canonical catalog entry.
Configuration
The server resolves auth in this order:
RUNAPI_API_KEYenvironment variable, useful for headless and CI hosts~/.config/runapi/config.json, created by the MCPlogintool orrunapi login- No key, which still allows
check_pricing
The config file is normally managed by login. A pre-provisioned headless config can use:
{
"apiKey": "your_runapi_key"
}
Do not commit real API keys.
Links
| Resource | URL |
|---|---|
| Wan model page | https://runapi.ai/models/wan |
| npm package | @runapi.ai/wan-mcp |
| GitHub repository | runapi-ai/wan-mcp |
| RunAPI MCP overview | runapi.ai/mcp |
| RunAPI docs | runapi.ai/docs |
License
Licensed under the Apache License, Version 2.0.
Wan MCP questions
How do I add Wan MCP to Claude Code?
Run claude mcp add --env 'RUNAPI_API_KEY=<YOUR_RUNAPI_API_KEY>' wan-mcp -- npx -y @runapi.ai/wan-mcp in your terminal. Add --scope user to make it available in every project.
How do I add Wan MCP to Cursor?
Open ~/.cursor/mcp.json and add the server under mcpServers, using the Cursor config shown in the install section above.
What tools does Wan MCP provide?
It exposes 8 tools, including animate, edit_video, image_to_video, speech_to_video, text_to_image.
How do I authenticate with the server?
You can use the RUNAPI_API_KEY environment variable or use the login tool to save credentials to ~/.config/runapi/config.json.
Can I check pricing without logging in?
Yes, the check_pricing tool works even without an active sign-in.
How many models are supported?
The server provides access to 18 different model variants across 6 endpoints.
Can I run tasks in the background?
Yes, you can instruct your assistant to create a task without waiting for it to finish by setting wait to false.