Luma MCP
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| TOP_P | No | 核采样阈值 | 0.95 |
| LUMA_DEBUG | No | `1`/`true` 时等同开启 `INCLUDE_META` | |
| MAX_TOKENS | No | 最大生成 token 数（部分模型有硬上限） | 8192 |
| MODEL_NAME | No | 模型名称覆盖（默认自动选择） | |
| MULTI_CROP | No | 大图多裁剪，设为 `false` 关闭 | true |
| CUSTOM_PATH | No | custom 提供商的 API 路径 | /chat/completions |
| TEMPERATURE | No | 采样温度 | 0.7 |
| INCLUDE_META | No | 为 `true` 时在工具结果末尾附加预处理/API 耗时等 meta | false |
| MCP_HTTP_HOST | No | HTTP 模式监听地址（Docker 内需为 `0.0.0.0`） | 0.0.0.0 |
| MCP_HTTP_PORT | No | HTTP 模式监听端口 | 3000 |
| MCP_TRANSPORT | No | 传输方式：`stdio`（默认）或 `http`（Streamable HTTP） | stdio |
| ZHIPU_API_KEY | No | Zhipu 提供商的 API Key | |
| CUSTOM_API_KEY | No | custom 提供商的 API Key | |
| MCP_HTTP_TOKEN | No | HTTP 模式 Bearer token；局域网共享务必设置 | |
| MODEL_PROVIDER | No | 模型提供商：`zhipu`、`siliconflow`、`qwen`、`volcengine`、`hunyuan`、`custom` | zhipu |
| CUSTOM_BASE_URL | No | custom 提供商的 OpenAI 兼容 API 地址 | https://api.minimaxi.com/v1 |
| ENABLE_THINKING | No | 思考模式，设为 `false` 关闭 | true |
| HUNYUAN_API_KEY | No | Hunyuan 提供商的 API Key | |
| CUSTOM_MODEL_NAME | No | custom 提供商的模型名称 | MiniMax-M3 |
| CUSTOM_TIMEOUT_MS | No | custom 提供商的超时毫秒 | 60000 |
| DASHSCOPE_API_KEY | No | Qwen 提供商的 API Key | |
| BASE_VISION_PROMPT | No | 自定义基础视觉提示词（设为空字符串可关闭） | |
| CUSTOM_AUTH_HEADER | No | custom 提供商的鉴权方式：`bearer` / `x-api-key` / `custom` | bearer |
| VOLCENGINE_API_KEY | No | Volcengine 提供商的 API Key | |
| SILICONFLOW_API_KEY | No | SiliconFlow 提供商的 API Key | |
| CUSTOM_THINKING_MODE | No | custom 提供商的思考模式：`disabled` / `openai` / `qwen_extra_body` | disabled |
| MULTI_CROP_MAX_TILES | No | 多裁剪最大图块数（含原图，1–16） | 5 |
| CUSTOM_AUTH_HEADER_VALUE | No | custom 提供商的自定义 Header 模板（`{{key}}` 会被替换为 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