Luma MCP
by Gupta5324
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| TOP_P | No | Nucleus sampling threshold. | 0.95 |
| LUMA_DEBUG | No | When set to `1` or `true`, behaves the same as enabling `INCLUDE_META`. | |
| MAX_TOKENS | No | Maximum number of generated tokens. Note: SiliconFlow (DeepSeek-OCR) is internally capped at 4096; setting very high values (e.g. 16384) may cause 400 errors on large images. | 8192 |
| MODEL_NAME | No | Overrides the model name. If unset, the provider's default model is selected automatically (e.g. `doubao-seed-1-6-vision-250815` for Volcengine, `hy-vision-2.0-instruct` for Hunyuan). | |
| MULTI_CROP | No | Enable multi-crop for large images; set to `false` to disable. | true |
| CUSTOM_PATH | No | API path for the `custom` provider. | /chat/completions |
| TEMPERATURE | No | Sampling temperature. | 0.7 |
| INCLUDE_META | No | When `true`, appends preprocessing/API timing metadata (`luma_meta`) to the end of the tool result. | false |
| MCP_HTTP_HOST | No | Listen address for HTTP mode (must be `0.0.0.0` inside Docker). | 0.0.0.0 |
| MCP_HTTP_PORT | No | Listen port for HTTP mode. | 3000 |
| MCP_TRANSPORT | No | Transport mode: `stdio` (default, local process) or `http` (Streamable HTTP). | stdio |
| ZHIPU_API_KEY | No | API key for the Zhipu provider (default model `glm-4.6v`). | |
| CUSTOM_API_KEY | No | API key for the `custom` provider, used with any OpenAI-compatible endpoint. | |
| MCP_HTTP_TOKEN | No | Bearer token for HTTP mode. Empty by default (no authentication). Strongly recommended when sharing the server on a LAN, since anyone able to reach the port can call `image_understand` and consume your model API quota. | |
| MODEL_PROVIDER | No | Model provider: `zhipu`, `siliconflow`, `qwen`, `volcengine`, `hunyuan`, or `custom`. | zhipu |
| CUSTOM_BASE_URL | No | OpenAI-compatible API base URL for the `custom` provider. | https://api.minimaxi.com/v1 |
| ENABLE_THINKING | No | Thinking mode; set to `false` to disable. | true |
| HUNYUAN_API_KEY | No | API key for the Hunyuan provider (default model `hunyuan-t1-vision-20250916`). | |
| CUSTOM_MODEL_NAME | No | Model name for the `custom` provider. | MiniMax-M3 |
| CUSTOM_TIMEOUT_MS | No | Request timeout in milliseconds for the `custom` provider. | 60000 |
| DASHSCOPE_API_KEY | No | API key for the Qwen provider (default model `qwen3-vl-flash`). | |
| BASE_VISION_PROMPT | No | Custom base vision prompt. Defaults to the built-in prompt; set to an empty string to disable. | |
| CUSTOM_AUTH_HEADER | No | Authentication header style for the `custom` provider: `bearer`, `x-api-key`, or `custom`. | bearer |
| VOLCENGINE_API_KEY | No | API key for the Volcengine provider (default model `doubao-seed-1-6-flash-250828`). | |
| SILICONFLOW_API_KEY | No | API key for the SiliconFlow provider (default model `deepseek-ai/DeepSeek-OCR`). | |
| CUSTOM_THINKING_MODE | No | Thinking mode for the `custom` provider: `disabled`, `openai`, or `qwen_extra_body`. | disabled |
| MULTI_CROP_MAX_TILES | No | Maximum number of tiles for multi-crop (including the original image), 1-16. | 5 |
| CUSTOM_AUTH_HEADER_VALUE | No | Custom header template for the `custom` provider, e.g. `"X-API-Key: {{key}}"`; `{{key}}` is replaced with the API key. |
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
Server capabilities have not been inspected yet.
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
No tools | |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
This server cannot be deployed
Maintenance
ActivityActive
ResponsivenessNo issues