mimo-mcp
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| MIMO_API_KEY | Yes | Your MiMo API key from platform.xiaomimimo.com | |
| MIMO_BASE_URL | No | Optional base URL for Token Plan users (e.g., https://token-plan-cn.xiaomimimo.com/v1) | |
| MIMO_DEFAULT_MAX_TOKENS | No | Default max tokens for thinking models (default: 4096) |
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
} |
| prompts | {
"listChanged": false
} |
| resources | {
"subscribe": false,
"listChanged": false
} |
| experimental | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| mimo.chatC | 多模态对话。messages 兼容 OpenAI 格式,可混入图像/视频。 |
| mimo.image_understandC | 图像理解。images 数组每项可选 url/path/base64。 |
| mimo.video_understandB | 视频理解。video 参数支持 4 种形式自动识别:(a) 直链 mp4 URL;(b) B 站/YouTube/抖音/小红书等视频站链接(yt-dlp 自动下载);(c) 本地路径(绝对/相对/~);(d) data:video/mp4;base64,... DataURL。 |
| mimo.ttsA | 文本合成语音。voice 选预置(mimo_default/冰糖/茉莉/苏打/白桦/Mia/Chloe/Milo/Dean)或自建 voice_id;instructions 可传自然语言风格指令(v2.5 导演模式:角色/场景/指导);文本里也可嵌 (风格)/[音频标签]/(唱歌) 标签。返回本地 wav 路径。 |
| mimo.voice_clone_createA | 上传参考音频创建克隆音色。立即跑一次试听验证。返回 voice_id 后,在 mimo.tts 里直接传 voice_id=<返回值> 即可用克隆音色朗读任意文本。 |
| mimo.voice_design_createB | 文字描述生成自定义音色 + 试听。MiMo voicedesign 是 stateless 的,本工具会把 prompt 入库以便复用,但每次朗读会基于 prompt 重新合成。 |
| mimo.voice_listB | 列出本地音色库(默认 + 克隆 + 设计)。 |
| mimo.voice_deleteC | 从本地音色库删除。 |
| mimo.asrA | 语音转写(F7)。传 audio_path(本地文件)或 audio_url(直链)之一,language 可选 auto / zh / en。默认走 mimo-v2.5-asr,返回纯文本。 |
| mimo.healthA | 健康检查:配置 / 网络 / 鉴权 / ASR 云端可用性。 |
| mimo.usageC | 本地 audit_log 聚合的最近用量。 |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 11 tools
Each tool targets a distinct modality or function (ASR, TTS, chat, image/video understanding, voice cloning, health check, usage). No two tools have overlapping purposes; the dedicated image and video understanding tools are separate from the multimodal chat, so an agent can easily distinguish them.
Tool names follow a 'mimo.<action>_<noun>' pattern only partially. Some are fully verb_noun (image_understand, voice_clone_create), while others are just nouns (mimo.asr, mimo.chat, mimo.tts). This mix reduces predictability, though the names remain readable and clear.
With 11 tools, the server covers a broad multimodal AI scope—speech, text, image, video, voice cloning, and system monitoring—without being overwhelming. Each tool serves a well-defined purpose, making the count appropriate for the domain.
The tool set provides comprehensive coverage for a multimodal interaction platform: audio transcription, text-to-speech, image/video understanding, voice cloning (create, list, delete), voice design, multimodal chat, health check, and usage monitoring. No obvious lifecycle gaps exist for the stated purpose.