A FastAPI-based REST service for Qwen3 text-to-speech optimized for consistent voice cloning and high-speed audio synthesis.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install chichi-speech
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 chichi-speech 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 Chichi Speech?
Chichi Speech is a specialized RESTful service designed to bring advanced text-to-speech (TTS) capabilities to your development environment. Built on top of the Qwen3 model library and FastAPI, this skill allows users to generate high-fidelity audio from text while leveraging sophisticated voice cloning technology. By utilizing a specific voice prompt, the service ensures that the cloned voice remains consistent across multiple generation requests, making it a valuable tool within the Openclaw Skills ecosystem.
This service is particularly effective for developers who need a reliable, self-hosted TTS backend that prioritizes performance. It minimizes re-computation by reusing reference audio prompts, ensuring that your AI agents or applications can produce speech quickly without sacrificing quality. Integrating this tool as one of your primary Openclaw Skills enables seamless voice interaction for various automated workflows.
Chichi Speech Use Cases
- Creating consistent voice-overs for digital content using a single reference audio file.
- Enhancing AI agents with realistic vocal identities through high-quality voice cloning.
- Building automated notification systems that require natural-sounding speech output.
- Expanding the capabilities of your Openclaw Skills library to include professional-grade speech synthesis.
How Chichi Speech Works
- The user initializes the service via the CLI, specifying a reference audio file and its corresponding text transcript.
- The FastAPI server starts locally, hosting a specialized /synthesize endpoint for audio generation.
- The skill pre-computes the voice prompt characteristics to optimize subsequent synthesis requests.
- When a POST request is received, the Qwen3 model processes the target text using the pre-loaded voice profile.
- The system generates and returns a high-quality WAV file, ready for immediate playback or integration.
Chichi Speech Setup
To get started with this addition to your Openclaw Skills, ensure you have Python 3.10 or higher. Install the package using pip:
pip install -e .
Launch the service by defining your host, port, and reference voice parameters:
chichi-speech --port 9090 --host 127.0.0.1 --ref-audio "[YOUR_AUDIO_URL]" --ref-text "[YOUR_REFERENCE_TEXT]"
Chichi Speech Data Schema & Taxonomy
The skill manages audio generation through structured API interactions and localized file outputs as detailed below:
| Component | Type | Description |
|---|---|---|
| Input Request | JSON | Accepts text and language parameters for synthesis. |
| Audio Output | WAV | Generates high-fidelity audio files based on the Qwen3 model. |
| Reference Data | URL/String | Utilizes audio and text prompts to define the cloning target. |
| API Docs | OpenAPI | Auto-generated FastAPI documentation available at the /docs path. |
Chichi Speech Advanced Features
- Specialized voice cloning optimization that avoids redundant computation for consistent voices.
- Full support for custom reference audio, allowing users to clone any voice provided via a URL.
- Highly configurable CLI allowing for customized port and host settings to fit any local or cloud environment.
- Streamlined integration for developers building complex workflows with Openclaw Skills.
SKILL.md
Loading
Related Openclaw Skills
Vehicle Expense Tracker
A multi-language vehicle expense management tool that synchronizes fuel and maintenance logs to Google Sheets or local Excel files.
News Fetcher (hfnews)
A specialized news aggregation tool for AI agents that fetches and filters technical news while stripping out political and sports noise.
Coala Client
A versatile command-line interface for interacting with LLMs and managing Model Context Protocol (MCP) servers and AI skills.
Research Report Generator
An automated research assistant that analyzes technical papers and codebases to generate comprehensive, multi-iteration reports with PDF export capabilities.
Ichiro-Mind
Ichiro-Mind is a sophisticated unified memory architecture that provides AI agents with persistent, hierarchical recall through a 4-layer system of hot, warm, cold, and archive storage.
Private Chat
An encrypted communication layer for AI agents that secures sensitive data using AES-256-CBC and stealth activation.
Featured*
