A sophisticated text-to-speech engine powered by the Xiaomi MiMo v2 model, enabling AI agents to generate expressive audio with emotional and dialectal nuance.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install xiaomimimotts
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 xiaomimimotts 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 MiMo TTS?
MiMo TTS is a high-performance speech synthesis skill designed for the Openclaw Skills ecosystem. It utilizes the advanced MiMo-v2-tts model to transform plain text into high-quality WAV audio files. Beyond simple conversion, it provides developers with deep control over vocal delivery, making it ideal for creating more human-like interactions within AI-driven applications.
The skill stands out by offering native support for multiple languages and specialized styles. Whether you need a standard professional tone, a specific regional dialect like Dongbei, or an emotional performance (e.g., happy, whispering, or even singing), MiMo TTS provides the technical hooks to achieve realistic prosody and expression through its unique audio tagging system.
MiMo TTS Use Cases
- Creating vocal responses for AI assistants and conversational agents.
- Generating localized content using specific regional dialects and accents.
- Producing expressive storytelling or role-play audio with embedded emotional cues.
- Automating the creation of voiceovers for notifications, alerts, or multimedia content.
- Enhancing accessibility by providing natural-sounding text-to-speech for long-form reading.
How MiMo TTS Works
- The user provides a text input along with optional parameters for voice, style, and output path.
- The Openclaw Skills runner validates the required MIMO_API_KEY environment variable.
- A Python-based synthesis script processes the text, interpreting any embedded style tags or audio tags.
- The request is sent to the Xiaomi MiMo API backend for high-fidelity processing.
- The system generates a WAV audio file at the specified destination, ready for playback or distribution.
MiMo TTS Setup
To use this within the Openclaw Skills framework, you must first configure your credentials:
- Obtain an API key from the Xiaomi MiMo developer portal.
- Configure the skill using the Openclaw CLI:
openclaw config set skills.entries.mimo-tts.apiKey "your-api-key-here"
Alternatively, export the environment variable directly:
export MIMO_API_KEY="your-api-key-here"
MiMo TTS Data Schema & Taxonomy
The skill manages audio generation through the following parameter schema and output structure:
Parameter Taxonomy
| Parameter | Type | Description |
|---|---|---|
| Text | String | The raw text or tagged content to be synthesized. |
Voice (-v) |
Enum | Selection of mimo_default, default_zh, or default_en. |
Style (-s) |
String | Controls emotion, dialect (e.g., Dongbei), or speed. |
Format (-f) |
String | Default is wav. |
Output Files
Generated files are saved locally as standard audio assets (e.g., output.wav).
MiMo TTS Advanced Features
- Granular Audio Tags: Use inline markers like
(deep breath),(bitter smile), or(silence)to insert non-verbal cues into the speech. - Style Encapsulation: Wrap text in
<style>blocks to apply global modifiers such as singing or specific character roles. - Dialect Support: Native capabilities for regional variations, allowing for more authentic localized user experiences.
- Dynamic Speed Control: Adjust the tempo of the speech output via style tags to match the urgency or context of the message.
- Multi-Voice Support: Toggle between optimized Chinese and English female voices for international Openclaw Skills deployments.
SKILL.md
Loading
Related Openclaw Skills
Hermes Learning Loop
A self-improving autonomous learning system for AI agents that extracts successful workflows into reusable skills and maintains a multi-layered memory architecture.
Memory Boost
A dual-purpose utility providing durable task documentation and an automated keep-alive monitoring loop for persistent AI agent workflows.
Role Configurator
A conversational setup tool that helps users define their AI assistant's persona, tasks, and communication style through guided prompts or templates.
osascript-email
A macOS-specific skill for sending emails via Mail.app using AppleScript, bypassing the need for SMTP credentials or API keys.
Tiered Recall
A sophisticated memory management system that overcomes LLM context limits by intelligently loading core project data, recent logs, and indexed memories.
Browser Use
A high-performance browser automation skill providing persistent sessions, low-latency interactions, and seamless Chrome profile integration.
Featured*
