A command-line AI chatbot with real-time streaming responses and text-to-speech playback.
- Streaming chat responses powered by Replicate
- Sentence-level text-to-speech using Piper TTS
- Persistent conversation history stored as JSON
- Verbose mode for detailed request and chunk logging
- Configurable system prompt and generation settings
git clone <repository-url>
cd Floura
python -m venv venv
source venv/bin/activate
pip install -r requirements.txt --break-system-packagesRefer to INSTALLATION.md for full setup instructions, including voice model downloads and environment configuration.
python src/main.pyAvailable commands:
| Command | Description |
|---|---|
/help |
Show available commands |
/verbose on / /verbose off |
Toggle detailed logging |
/clear |
Clear conversation history |
/history |
Display full conversation history |
/system |
Edit the system prompt |
/settings |
Show current configuration |
/exit |
Exit the application |
src/
main.py
services/
chat_service.py
tts_service.py
utils/
chat_utils.py
data/
voices/
audio/
history.json
Internal project. No license specified.