vidu

Vidu MCP Server exposes a set of tools to generate videos from images using Vidu's AI models. The server supports converting a single image into a video with configurable duration, resolution, model (viduq1, vidu1.5, vidu2.0), movement amplitude, and optional background music. It also provides a way to upload images for use with the Vidu API and check the status of ongoing generation tasks, including credit usage and progress. You can configure a callback URL to receive asynchronous notifications when the generation state changes. The server is designed to be invoked via the MCP interface exposed in Gemini CLI, with the actual server script executed through Node, and requires a Vidu API key to authenticate API requests. In practice, you build the TypeScript sources, load the MCP into Gemini CLI, and then issue requests for image-to-video conversion, status checks, or image uploads through the MCP endpoints.

Prerequisites:

• Node.js v14 or higher
• npm (comes with Node.js)
• A Vidu API key
• TypeScript development tooling (optional for local build)

Manual installation steps:

1. Clone the repository

git clone https://github.com/el-el-san/vidu-mcp-server.git
cd vidu-mcp-server

2. Install dependencies

npm install

3. Create an environment file with your Vidu API key (or provide via MCP env):

# .env (example)
VIDU_API_KEY=your_api_key_here

4. Build the TypeScript sources (if applicable for your setup)

npm run build

5. Run or expose the MCP server via your preferred method (Node.js execution per the mcp_config). For Gemini CLI usage, ensure the built entry point exists at the path you reference (build/index.js in this example).

Notes and tips:

• Ensure the Vidu_API_KEY is valid and has the required permissions for image-to-video generation and image uploads.
• When using Gemini CLI, update the path in the MCP config to the actual build/index.js location after building.
• The MCP supports multiple models (viduq1, vidu1.5, vidu2.0) with model-specific constraints on duration and resolution.
• For 4-second videos, some models may have default resolutions; adjust the resolution accordingly (e.g., 360p, 720p, 1080p).
• If you encounter upload or network errors, verify that the server has outbound internet access and the Vidu API endpoint is reachable.
• Use the callback_url parameter to receive asynchronous task updates; ensure your endpoint can handle POST notifications.
• In Gemini CLI settings, set "disabled": false to enable the server and restart Gemini after configuration changes.