README.md

---
title: Video Processor API
emoji: 🎬
colorFrom: blue
colorTo: indigo
sdk: docker
app_port: 7860
app_file: main.py
pinned: false
---

# 🎧 Video Processor API (Whisper + Gemini)

FastAPI-based video processor that:
- Extracts audio from video using FFmpeg
- Transcribes with Faster Whisper
- Summarizes using Google Gemini API
- Returns transcript, summary, and subtitles

### 🔑 Environment Variables
| Name | Description |
|------|--------------|
| `GEMINI_API_KEY` | Your Google Generative AI API key |
| `WHISPER_MODEL` | Whisper model size (e.g., tiny, base, small) |
| `WHISPER_DEVICE` | cpu or cuda |
| `MAX_UPLOAD_MB` | Max upload file size |

### 🚀 Run Locally
```bash
pip install -r requirements.txt
python app.py