Vidu Model Context Protocol Server

This is the Model Context Protocol (MCP) gateway designed for interfacing with the Vidu video generation API service. It furnishes utilities for synthesizing motion pictures from still frames utilizing Vidu's potent AI architecture.

Core Capabilities

• Still-to-Motion Conversion: Fabricate video clips from static visuals with adaptable parameters.
• Multi-Model Support: Compatible with viduq1, vidu1.5, and vidu2.0 iterations
• Model-Specific Constraints: Adherence to temporal duration and resolution mandates per model
• Auditory Accompaniment: Optional background music track support for 4-second outputs
• Asynchronous Feedback: Configuration for callback URLs to signal completion status changes
• Progress Tracking: Monitor the status of video generation jobs, including associated credit consumption metrics.
• Asset Ingestion: Streamlined procedure for uploading imagery destined for use by the Vidu backend (10MB size constraint).

Prerequisites

• Runtime Environment: Node.js (version 14 or newer)
• Authorization Credential: A valid Vidu API Key (obtainable from the Vidu official site)
• Development Toolchain: TypeScript (required for building/development)

Installation Procedures

Deployment via Smithery

Automated setup of the Vidu Video Synthesis Server for use within Claude Desktop environments utilizing Smithery:

bash npx -y @smithery/cli install @el-el-san/vidu-mcp-server --client claude

Gemini CLI Configuration

To integrate this server within the Gemini Command Line Interface environment, append the following server configuration details to your ~/.gemini/settings.json file:

{ "mcpServers": { "vidu": { "command": "node", "args": [ "your_path/vidu-mcp-server/build/index.js" ], "env": { "VIDU_API_KEY": "your_api_key_here" } } } }

Crucial Note: Substitute your_path with the actual directory where the software resides, and replace your_api_key_here with your authentic Vidu secret key.

Manual Software Acquisition

1. Clone this repository source code: bash git clone https://github.com/el-el-san/vidu-mcp-server.git cd vidu-mcp-server

2. Install necessary dependencies: bash npm install

3. Establish the configuration file by copying .env.template to .env and injecting your Vidu access credential:

VIDU_API_KEY=your_api_key_here

Operational Guide

For Gemini CLI Integration

1. Compile the TypeScript source code into executable JavaScript: bash npm run build

2. Apply the configuration specified in the Gemini CLI Setup section above.

3. Restart the Gemini CLI application to ensure the new MCP service is loaded.

Available Utilities (Tools)

1. Still Image to Motion Picture Synthesis

Converts a source still image into a video sequence based on user-defined attributes.

Parameters: - image_url (Mandatory): The publicly accessible Uniform Resource Locator pointing to the source image. - prompt (Optional): Descriptive text input guiding the video generation process (Maximum 1500 characters). - duration (Optional): Target length of the resultant video clip (in seconds) (Model dependent). - viduq1: Restricted exclusively to a 5-second duration. - vidu1.5/vidu2.0: Capable of 4 or 8 seconds (4 seconds is the default setting). - model (Optional): Designation of the generative AI architecture to employ ("viduq1", "vidu1.5", "vidu2.0"; defaults to "vidu2.0"). - resolution (Optional): The spatial dimensions of the output video (constrained by model and duration). - viduq1 (5s): Only supports 1080p definition. - vidu1.5/vidu2.0 (4s): Acceptable values are "360p", "720p", "1080p" (Default: "360p"). - vidu1.5/vidu2.0 (8s): Exclusively supports "720p" resolution. - movement_amplitude (Optional): Control over the extent of object motion within the rendered frames ("auto", "small", "medium", "large"; default is "auto"). - seed (Optional): An integer value for reproducible generation results. - bgm (Optional): Boolean flag to include background musical score (Applies only to 4-second renderings; default false). - callback_url (Optional): A designated endpoint to receive asynchronous notifications upon status transitions.

Example Request:

{ "image_url": "https://example.com/image.jpg", "prompt": "A tranquil lake scene set against mountain scenery", "duration": 8, "model": "vidu2.0", "resolution": "720p", "movement_amplitude": "medium", "seed": 12345, "bgm": false }

2. Synthesis Task Status Inquiry

Retrieves the current processing state for any active video generation job.

Parameters: - task_id (Mandatory): The unique identifier returned from the image-to-video synthesis utility.

Example Request:

{ "task_id": "12345abcde" }

3. Source Image Upload

Transfers a local image file to the Vidu infrastructure for subsequent processing.

Parameters: - image_path (Mandatory): The absolute or relative file system location of the image asset. - image_type (Mandatory): The format extension of the file ("png", "webp", "jpeg", "jpg").

Example Request:

{ "image_path": "/path/to/your/image.jpg", "image_type": "jpg" }

Troubleshooting Guidance

• Credential Authentication Failures: Verify that the Vidu API Key is correctly placed in either the .env configuration file (manual mode) or the Gemini CLI settings object (CLI mode).
• Asset Transfer Errors: Confirm that the image file is valid and adheres to size restrictions (10MB limit for the dedicated upload tool; remote URLs support up to 50MB).
• Network Interruption: Ensure active internet connectivity allows successful communication with the Vidu API endpoints.
• Gemini CLI Integration Issues:
• Confirm that the server compilation (npm run build) completed successfully prior to CLI setup.
• Double-check that the path specified in settings.json accurately points to the build/index.js entry point.
• Remember to relaunch the Gemini CLI application after modifying configuration files.
• Ensure the server configuration block does not contain "disabled": true (it should be false or omitted).