ChatterBox-FastAPI

High-performance Text-to-Speech server with OpenAI-compatible API

Key ChatterBox Details

• SoTA zeroshot TTS
• 0.5B Llama backbone
• Unique exaggeration/intensity control
• Ultra-stable with alignment-informed inference
• Trained on 0.5M hours of cleaned data
• Watermarked outputs
• Easy voice conversion script
• Outperforms ElevenLabs

Tips

• General Use (TTS and Voice Agents):

  • The default settings (exaggeration=0.5, cfg_weight=0.5) work well for most prompts.
  • If the reference speaker has a fast speaking style, lowering cfg_weight to around 0.3 can improve pacing.

• Expressive or Dramatic Speech:

  • Try lower cfg_weight values (e.g. ~0.3) and increase exaggeration to around 0.7 or higher.
  • Higher exaggeration tends to speed up speech; reducing cfg_weight helps compensate with slower, more deliberate pacing.

Features

• OpenAI API Compatible: Drop-in replacement for OpenAI's /v1/audio/speech endpoint

• Custom Voice Cloning: '/custom_voice' UI Generate, Sample and save custom voice for reuse in API Calls

• Smooth Transitions: Crossfaded audio segments for seamless listening experience

  (Route implemented UI Coming Soon)

• TODO: Web UI Configuration: Configure all server settings directly from the interface

• TODO: Dynamic Environment Variables: Update API endpoint, timeouts, and model parameters without editing files

• TODO: Server Restart: Apply configuration changes with one-click server restart

Getting Started

Run without Docker:

1. (Recommended) Create and activate a virtual environment:

   git clone https://github.com/getfit-us/ChatterBox-FastAPI.git &&
   cd ChatterBox-FastAPI 
   python3 -m venv .venv
   source .venv/bin/activate

2. Install dependencies:

   pip install --upgrade pip
   pip install -r requirements.txt
   cp .env.example .env

3. Run the FastAPI server:

   
   python app.py
   

4. The API will be available at http://localhost:8880/docs

Run with Docker Compose:

1. Build and start the server:

   cd docker
   docker-compose up --build

2. The API will be available at http://localhost:8880/docs

3. To stop the server:

   docker-compose down

Run with Docker only (no Compose):

1. Build the image:

   docker build -t chatterbox-fastapi -f docker/Dockerfile .

2. Run the container:

   docker run --rm -p 8880:8880 \
     -v $(pwd)/outputs:/app/outputs \
     -v $(pwd)/voices:/app/voices \
     -v $(pwd)/static:/app/static \
     --env-file .env \
     chatterbox-fastapi