Chrome extension + Python server for realtime English→Vietnamese meeting translation on Microsoft Teams (web).
Hỗ trợ OpenAI GPT-4o-mini và DeepSeek v4-flash.
- Chrome Extension captures Teams tab audio via
chrome.tabCapture - Local Python server runs Whisper STT on RTX 5080 GPU
- OpenAI GPT-4o-mini / DeepSeek v4-flash API handles translation
- Subtitles injected as DOM overlay
- Install dependencies:
pip install -r requirements.txt - Start server:
python server/server.py - Configure API key at
http://127.0.0.1:8765/admin - Load extension:
chrome://extensions→ "Load unpacked" → selectextension/ - Open Teams meeting → click extension icon → Start
- Python 3.12+
- NVIDIA GPU with CUDA 12+ (RTX 5080 recommended)
- Chrome 120+
- OpenAI API key or DeepSeek API key
VocalAgent/
├── server/ # Python FastAPI backend
│ ├── config.py # ConfigManager with defaults
│ ├── pipeline.py # VAD, WhisperSTT, Translator, Pipeline
│ └── server.py # FastAPI app (/admin, /ws)
├── extension/ # Chrome extension (Manifest V3)
│ ├── manifest.json
│ ├── popup.html/js/css
│ ├── background.js # Service worker
│ ├── content.js # Audio capture + subtitle overlay
│ └── overlay.css
├── test/ # Python tests
│ ├── test_config.py
│ ├── test_translator.py
│ ├── test_vad.py
│ ├── test_whisper.py
│ ├── test_ws.py
│ ├── test_pipeline.py
│ └── fixtures/
└── requirements.txt
pip install -r requirements.txt
python -m pytest test/ -v- Go to
chrome://extensions - Enable "Developer mode"
- Click "Load unpacked" → select
extension/folder - Open Teams meeting → click extension icon → Start
Cách 1 — Environment variables (ưu tiên cao nhất):
Tạo file .env (xem .env.example):
VOCALAGENT_PROVIDER=openai
VOCALAGENT_OPENAI_API_KEY=...
# hoặc VOCALAGENT_DEEPSEEK_API_KEY=...
Cách 2 — Admin UI:
http://127.0.0.1:8765/admin — chọn provider + nhập key.
Cách 3 — Config file:
config.json (auto-created, auto-migrated to SQLite).
| Provider | Default Model | Env Var |
|---|---|---|
| OpenAI | gpt-4o-mini |
VOCALAGENT_OPENAI_API_KEY |
| DeepSeek | deepseek-v4-flash |
VOCALAGENT_DEEPSEEK_API_KEY |