mcp-video-gen
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| VEO_MODEL | No | Optional. Default: veo-3.1-fast-generate-001 | |
| GCP_REGION | No | Optional. Default: us-central1 | |
| ARK_API_KEY | No | Volcengine Ark API key. At least one provider API key is required. | |
| VIDU_API_KEY | No | Vidu API key. | |
| GCP_PROJECT_ID | No | Google Cloud Project ID. Required for Google Veo, Lyria, and STT. | |
| GEMINI_API_KEY | No | Google API key for Veo. Recommended. Alternative uses ADC. | |
| VEO_GCS_BUCKET | No | Optional. GCS bucket for output. If unset, uses base64 inline. | |
| ARK_VIDEO_MODEL | No | Optional. Default: doubao-seedance-2-0-fast-260128 | |
| MINIMAX_API_KEY | No | MiniMax API key. Enables video, TTS, and music generation. | |
| KLING_ACCESS_KEY | No | Kling AI access key. | |
| KLING_SECRET_KEY | No | Kling AI secret key. | |
| MINIMAX_API_HOST | No | Optional. Default: https://api.minimax.chat | |
| VIDEO_OUTPUT_DIR | No | Optional. Default: ./output | |
| ARK_VIDEO_API_KEY | No | Volcengine Ark video-specific API key. Overrides ARK_API_KEY for video generation. | |
| DASHSCOPE_API_KEY | No | DashScope API key for Wan video generation. | |
| ARK_VIDEO_BASE_URL | No | Optional. Default: https://ark.cn-beijing.volces.com/api/v3 | |
| SILICONFLOW_API_KEY | No | SiliconFlow API key. | |
| ARK_VIDEO_RESOLUTION | No | Optional. Default: 720p | |
| DEFAULT_VIDEO_PROVIDER | No | Optional. Default prefers ark when configured. |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": false
} |
| resources | {
"subscribe": false,
"listChanged": false
} |
| experimental | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| generate_videoA | Generate a video from a text prompt (text-to-video) or from an image + prompt (image-to-video, ark/veo). Available providers: none configured. Default: none. |
| query_video_statusB | Query the status of a video generation task and download the result. |
| list_providersA | List all available video, TTS, and music providers. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
| Model Catalog | No providers configured. Set API keys to enable providers. |
TDQS
Scored across 3 tools
Each tool has a distinct, non-overlapping purpose: generating a video, querying its status, and listing providers. An agent can easily select the correct tool based on the task at hand.
All tool names follow a consistent verb_noun pattern in snake_case: generate_video, query_video_status, list_providers. The naming convention is uniform and predictable.
With only 3 tools, the set is tightly scoped to the core operations of video generation: create, monitor, and discover providers. Each tool is essential and earns its place without redundancy.
The lifecycle is largely covered: generation is initiated, status can be polled, and results downloaded. However, there is no explicit way to cancel or list past tasks, which is a minor gap but workable.