Dockerfile

FROM python:3.11-slim

WORKDIR /app

# Install CPU-only torch first (saves ~1.8GB vs CUDA build)
RUN pip install --no-cache-dir torch --index-url https://download.pytorch.org/whl/cpu

COPY requirements.txt .
RUN pip install --no-cache-dir -r requirements.txt

COPY app.py .
COPY download_models.py .

# Pre-download commonly used TTS models so first requests are fast.
RUN python download_models.py

EXPOSE 7860

CMD ["uvicorn", "app:app", "--host", "0.0.0.0", "--port", "7860"]