BridgeMate

This is a Flask application that provides an AI‑powered chat interface with speech‑to‑text and text‑to‑speech features. The frontend uses vanilla JavaScript and interacts with the backend over simple /api/* endpoints.

⚠️ Important The web UI must be loaded via the Flask server (e.g. http://localhost:4000). Opening index.html directly from the file system will lead to errors such as "Invalid URL" when you try to record audio, because the browser cannot resolve the relative API paths.

Running

uv venv .venv
source .venv/bin/activate   # or `.venv\\Scripts\\activate` on Windows
uv sync

# Configure environment variables
cp .env.example .env

# Start the app
uv run src/main.py

Environment variables

The app loads .env automatically at startup.

Required:

• FEATHERLESS_API_KEY
• HF_API_TOKEN
• ELEVENLABS_API_KEY

Optional:

• STT_PROVIDER (elevenlabs or whisper, default elevenlabs)

Use .env.example as the template for your local .env file.

Speech-to-text provider toggle

/api/speech-to-text supports two providers:

• elevenlabs (default): Uses ElevenLabs scribe_v2
• whisper: Uses Hugging Face Inference API model openai/whisper-large-v3

Choose the backend provider with the STT_PROVIDER environment variable before starting the server.

Then open your browser at http://localhost:4000 and start chatting or recording.