Quick Start
Choose a quick-start path based on your TTS provider preference.
Option 1: Cloud TTS (Easiest)
No Docker required. Use OpenAI, AWS Polly, or Google Cloud TTS.
1. Install Audify
pip install audify-cli
2. Configure credentials
Create a .keys file in your working directory:
# OpenAI TTS (simplest cloud option)
OPENAI_API_KEY=sk-your-openai-api-key
TTS_PROVIDER=openai
3. Convert a book
audify book.epub --task direct --tts-provider openai
Option 2: Kokoro TTS (Local, Free)
Requires Docker and a CUDA GPU. Fast and low-latency synthesis.
1. Clone and start services
git clone https://github.com/garciadias/audify.git
cd audify
docker compose --profile kokoro --profile ollama up -d
# Wait ~2-3 minutes for services to start
2. Pull an Ollama model
docker compose exec ollama ollama pull qwen3:30b
3. Convert a book
uv sync
audify book.epub --task direct
Option 3: Qwen-TTS (Local, Free)
Requires a local Qwen-TTS-compatible API and typically a GPU.
Option 3A: Docker Compose Profile (Recommended)
Run the provided qwen-tts service in this repository:
Start Qwen-TTS and Ollama:
docker compose --profile qwen up -d qwen-tts ollama
Confirm Qwen-TTS health:
curl http://localhost:8890/healthUse Qwen in Audify:
# Direct TTS audify book.epub --task direct --tts-provider qwen # Audiobook generation with Ollama LLM + Qwen-TTS audify book.epub --task audiobook --tts-provider qwen -m gemma4:31b
Option 3B: Local API Wrapper Script
Install dependencies:
pip install qwen-tts fastapi uvicorn soundfile numpy torch
Run the wrapper:
python scripts/qwen_tts_api.pyConfigure Audify:
TTS_PROVIDER=qwen QWEN_API_URL=http://localhost:8890 QWEN_TTS_VOICE=Vivian
Equivalent commands
audify book.epub --task direct --tts-provider qwen
audify book.epub --task audiobook --tts-provider qwen --llm-model gemma4:31b
Next Steps
Learn about all usage options including translation
Explore the task system for LLM-powered audio generation
Set up cloud LLM APIs for advanced audiobook creation