MiMo TTS MCP Server
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| MIMO_API_KEY | No | MiMo API Key, created at https://platform.xiaomimimo.com/console/api-keys. Required when calling the TTS tools. `XIAOMIMIMO_API_KEY` is accepted as an alternative variable name. The key is only read from the environment (never from a config file, never written to disk, never logged). | |
| MIMO_BASE_URL | No | Override the API base URL, e.g. when using a Token Plan or a self-hosted gateway. | https://api.xiaomimimo.com/v1 |
| XIAOMIMIMO_API_KEY | No | Alternative variable name for the MiMo API Key. Either this or `MIMO_API_KEY` must be set when calling the TTS tools. | |
| MIMO_TTS_OUTPUT_DIR | No | Directory where generated audio files are written. This variable also defines the write sandbox boundary, so it is recommended to set it per project. Supports `~` expansion; relative paths are resolved against the process cwd and converted to absolute paths. Paths outside this directory, `..` traversal, and symlinked subdirectories pointing outside are rejected. | ~/mimo-tts |
| MIMO_TTS_TIMEOUT_MS | No | Timeout in milliseconds for a single TTS request. | 180000 |
| MIMO_TTS_ALLOW_ANY_PATH | No | Set to `1` to allow writing to arbitrary paths outside of `MIMO_TTS_OUTPUT_DIR`. Leave unset to keep the default sandbox restriction. | 未设置 |
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": true
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| text_to_speechA | Synthesize speech with Xiaomi MiMo |
| design_voice_to_speechA | Synthesize speech with Xiaomi MiMo |
| clone_voice_to_speechA | Synthesize speech with Xiaomi MiMo |
| list_voicesA | List the MiMo TTS models, their capabilities, built-in preset voices, and where files are written. Read-only; no API key needed. |
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 4 tools
The three synthesis tools are cleanly differentiated by voice source: preset built-in voice (text_to_speech), invented voice from a description (design_voice_to_speech), and replicated voice from an audio sample (clone_voice_to_speech). list_voices is clearly read-only and metadata-focused, so there is no realistic misselection.
All three synthesis tools follow a predictable '<source>_to_speech' pattern, and the read-only tool uses the conventional 'list_<noun>' form. The convention is consistent and self-explanatory across the whole set.
Four tools is well-scoped for a TTS server: three synthesis modes plus one discovery/introspection tool. Each tool earns its place with no filler or redundancy.
The surface covers the key TTS lifecycle modes (preset, design, clone) plus model/voice discovery, which is strong for the domain. Minor gaps exist around managing generated output (e.g. listing or deleting written files), but agents can work around this via the returned file paths.