A local proxy service that converts OpenAI TTS API requests into Xiaomi MiMo platform calls with integrated FFmpeg transcoding.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install mimo-tts-proxy
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install mimo-tts-proxy using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Xiaomi MiMo TTS Proxy?
The Xiaomi MiMo TTS Proxy is a specialized extension within the Openclaw Skills ecosystem designed to bridge the gap between the OpenAI Text-to-Speech (TTS) API standard and the Xiaomi MiMo large model platform. By running a local HTTP service, it allows developers to use Xiaomi's high-quality speech synthesis in any application that supports the OpenAI TTS format, providing a cost-effective and flexible alternative for voice-enabled AI agents.
This skill is particularly powerful because it handles the heavy lifting of audio transcoding locally. Using FFmpeg, it can convert the raw Xiaomi output into various formats like Opus, MP3, and AAC, ensuring compatibility with platforms like Feishu, WeChat, or custom web interfaces. It functions as a lightweight, reliable middleware that simplifies the integration of regional AI voice services into global developer workflows.
Xiaomi MiMo TTS Proxy Use Cases
- Equipping Openclaw agents with natural-sounding Chinese and English voice responses.
- Building local TTS proxy services for applications that only support the OpenAI API schema.
- Transcoding Xiaomi MiMo audio on-the-fly into specialized formats like Opus for instant messaging platforms.
- Implementing emotional or stylized speech synthesis using custom style tags in automated workflows.
How Xiaomi MiMo TTS Proxy Works
- The proxy starts a local HTTP server (defaulting to port 18899) and waits for incoming POST requests at the /audio/speech endpoint.
- When a request arrives, the proxy maps the OpenAI-formatted payload (input text, voice selection) to the Xiaomi MiMo API requirements.
- The proxy forwards the request to the Xiaomi MiMo platform (api.xiaomimimo.com) using the configured API Key.
- Upon receiving the audio data from Xiaomi, the proxy utilizes local FFmpeg to transcode the stream into the user's requested format (e.g., MP3 or Opus).
- The final processed audio is returned to the client as a standard binary response.
Xiaomi MiMo TTS Proxy Setup
Prerequisites
- FFmpeg must be installed and accessible in your system PATH.
- Node.js ≥ 18 for native fetch support.
- A valid Xiaomi MiMo API Key.
Installation Steps
- Create the configuration file at
~/.openclaw/tts-proxy.env:
MIMO_TTS_BASE=https://api.xiaomimimo.com
MIMO_TTS_KEY=your_xiaomi_api_key
TTS_PROXY_PORT=18899
- (Recommended) Install and start the systemd service for persistence:
sudo cp mimo-tts-proxy.service /etc/systemd/system/
sudo systemctl daemon-reload
sudo systemctl enable --now mimo-tts-proxy.service
- Verify the service status:
curl http://127.0.0.1:18899/health
Xiaomi MiMo TTS Proxy Data Schema & Taxonomy
File Organization
| File | Description |
|---|---|
~/.openclaw/tts-proxy.mjs |
The core proxy logic script. |
~/.openclaw/tts-proxy.env |
Environment variables containing API keys and port settings. |
/etc/systemd/system/mimo-tts-proxy.service |
The service definition for background execution. |
API Parameters
| Parameter | Type | Default | Description |
|---|---|---|---|
| input | string | Required | The text to synthesize, supports <style> tags. |
| voice | string | mimo_default | Predefined voices: default_zh, default_en, mimo_default. |
| response_format | string | mp3 | Output formats: opus, mp3, aac, flac, wav, pcm. |
Xiaomi MiMo TTS Proxy Advanced Features
- Advanced Style Injection: Use
<style>happy</style>or<style>sad</style>tags at the beginning of the input text to control the emotional tone of the generated speech. - Dialect Support: Access regional accents including Cantonese, Sichuanese, and Northeastern Chinese via style tags.
- Multi-Format Transcoding: Real-time hardware-accelerated conversion using FFmpeg for optimized streaming.
- Systemd Integration: Full lifecycle management including auto-restart on failure and boot-time activation.
- Health Monitoring: Dedicated
/healthendpoint for integration with uptime monitors and Openclaw Skills health checks.
SKILL.md
Loading
Related Openclaw Skills
Video Maker AI Free
An advanced AI skill that transforms scripts, images, and raw ideas into polished, production-ready videos without manual editing.
Umeng Mini-Program Campaign Manager
A command-line skill for managing Umeng Mini-Program promotion links, campaigns, and channel assets.
Free AI Video Creator
A powerful AI-driven skill that converts text scripts, images, and raw ideas into structured, professional video content without production costs.
Emergence Diagram Rendering
A high-fidelity diagram generation skill enabling AI agents to produce Mermaid, D2, and Graphviz visuals with local-first rendering and automated self-correction.
Wyckoff Stock Diagnostic Tool
A sophisticated A-Share stock diagnostic tool implementing the Wyckoff 2.0 methodology for automated technical analysis and structural identification.
Database Migration Manager
A comprehensive tool for managing database lifecycles and migrations within the Openclaw Skills ecosystem.
Featured*
