An enhanced Text-to-Speech skill for Openclaw allowing multi-agent unique voice identities and native voice messaging for Telegram and Feishu.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install voice-for-openclaw
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 voice-for-openclaw 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 MiniMax TTS Plus?
MiniMax TTS Plus is a robust extension for the Openclaw ecosystem designed to provide high-fidelity voice generation. It allows each AI agent to maintain a distinct persona by selecting unique voice identities from MiniMax's library, including warm, sweet, or youthful tones. By integrating this into Openclaw Skills, developers can bridge the gap between text-based AI and immersive audio communication.
The skill is built with flexibility in mind, supporting native voice bubble delivery across multiple platforms. It handles the technical heavy lifting of audio transcoding—such as converting MP3 to OGG/Opus for Feishu—ensuring that voice messages appear as native, playable bubbles rather than just file attachments. This makes it an essential tool for developers building interactive assistants or notification bots.
MiniMax TTS Plus Use Cases
- Creating personalized AI assistants with unique voice identities in SOUL.md.
- Sending native, playable voice messages directly to Telegram chats.
- Delivering high-quality OGG/Opus voice bubbles to Feishu (Lark) users.
- Generating high-fidelity MP3 audio files for local use or secondary processing.
- Implementing multi-lingual voice responses for diverse user bases.
How MiniMax TTS Plus Works
- The user or agent triggers the skill by providing text and optional parameters like voice ID or model.
- The skill sends the request to the official MiniMax API endpoint (api.minimaxi.com).
- MiniMax generates the raw audio in MP3 format based on the selected model (e.g., speech-2.8-hd).
- For Telegram, the system sends the MP3 directly using the Bot API sendVoice method.
- For Feishu, the skill utilizes FFmpeg to transcode the MP3 into OGG/Opus format in approximately 25ms.
- The final audio asset is delivered to the target channel or saved locally depending on the command flags.
MiniMax TTS Plus Setup
Prerequisites
Ensure you have python3, ffmpeg, and the requests library installed on your system.
Environment Configuration
- Obtain your API Secret Key from the MiniMax Platform.
- Set up your environment variables in a
.envfile:
# Required
MINIMAX_API_KEY=your_secret_key_here
# Optional for channel delivery
TELEGRAM_BOT_TOKEN=your_bot_token
TELEGRAM_TARGET=chat_id
Installation
# Install FFmpeg (macOS)
brew install ffmpeg
# Install FFmpeg (Ubuntu/Debian)
sudo apt install ffmpeg
MiniMax TTS Plus Data Schema & Taxonomy
MiniMax TTS Plus organizes its data through structured API responses and local file management:
| Data Type | format | Description |
|---|---|---|
| Native Audio | MP3 | Default output for general use and Telegram. |
| Transcoded Audio | OGG/Opus | Generated specifically for Feishu native voice bubbles. |
| Voice Metadata | JSON | List of available system, cloned, and generated voices. |
| Configuration | .env | Secure storage for API keys and bot tokens. |
Each agent's identity is defined in their respective SOUL.md under the ## Voice Identity section, specifying the model, voice ID, and script path.
MiniMax TTS Plus Advanced Features
- Multi-agent support: Assign unique voices to different agents within the same environment.
- Auto-transcoding: Intelligent FFmpeg integration for platform-specific audio standards.
- Voice Discovery: Built-in
--list-voicesflag to query and categorize all available MiniMax voices. - Headless Generation: Use the
--generate-onlyflag for pure file creation without social channel delivery. - Quality Control: Choice between
speech-2.8-hdfor quality andspeech-2.8-turbofor speed.
SKILL.md
Loading
Related Openclaw Skills
ClawLite Video Content Engine
An intelligent content repurposing engine that transforms YouTube educational videos into comprehensive, multi-channel marketing packs and ClawLite-compatible assets.
GraphHopper
A powerful route planning and geocoding skill that enables AI agents to calculate travel paths and manage geographic data.
Zhuanzhuan Recycle Estimator
An expert AI recycling consultant that provides real-time market valuations for used electronics and appliances using text and image recognition.
Sure API Skill
A professional-grade interface for the Sure REST API that enables automated management of accounts, transactions, holdings, and valuations.
API Monitoring
An AI-driven framework for designing robust health check systems and alerting strategies for e-commerce API integrations.
Data Research Starter
A structured research initiation skill that converts ambiguous analysis requests into clear, verified research outlines through guided dialogue.
Featured*
