WhisperX API

GPU-powered transcription API in one command

Features • Quick Start • API • Config • Development

Turn any audio into text with speaker labels. No cloud. No limits. Just run:

uvx whisperx-api

WhisperX API wraps WhisperX in a REST API with speaker diarization, word-level timestamps, and multiple export formats. Self-hosted alternative to AssemblyAI, Deepgram, and Rev.ai.

Features

• Speaker Diarization - Identify who said what with pyannote
• Word-Level Timestamps - Precise alignment for every word
• Multiple Export Formats - SRT, WebVTT, TXT, JSON
• Webhook Callbacks - Get notified when transcription completes
• GPU Model Caching - Fast subsequent transcriptions
• Background Processing - Non-blocking async jobs
• Progress Tracking - Poll for real-time status

Quick Start

Prerequisites

• NVIDIA GPU with 6GB+ VRAM (or CPU mode for testing)
• CUDA 12.x drivers installed

Option A: One-Liner Install (Recommended)

curl -fsSL https://raw.githubusercontent.com/namastexlabs/whisperx-api/main/get-whisperx.sh | bash

This installs Python 3.12, uv, checks CUDA, and sets up whisperx-api.

Option B: Direct Run (if dependencies met)

uvx whisperx-api

Option C: pip install

pip install whisperx-api
whisperx-api

The API starts at http://localhost:8880. Swagger docs at /docs.

First Transcription

# Default API key is "namastex888" - works out of the box
curl -X POST http://localhost:8880/v1/transcript \
  -H "Authorization: namastex888" \
  -F "file=@audio.mp3"

# Check status (replace {id} with returned transcript ID)
curl http://localhost:8880/v1/transcript/{id} \
  -H "Authorization: namastex888"

API Reference

Submit Transcription

File upload:

curl -X POST http://localhost:8880/v1/transcript \
  -H "Authorization: namastex888" \
  -F "file=@audio.mp3"

URL download:

curl -X POST http://localhost:8880/v1/transcript \
  -H "Authorization: namastex888" \
  -F "audio_url=https://example.com/audio.mp3"

With speaker diarization:

curl -X POST http://localhost:8880/v1/transcript \
  -H "Authorization: namastex888" \
  -F "file=@audio.mp3" \
  -F "speaker_labels=true" \
  -F "speakers_expected=2"

Response Format

{
  "id": "550e8400-e29b-41d4-a716-446655440000",
  "status": "completed",
  "text": "Hello world, this is a transcription.",
  "words": [
    {"text": "Hello", "start": 0, "end": 500, "confidence": 0.98, "speaker": "A"}
  ],
  "utterances": [
    {"speaker": "A", "text": "Hello world...", "start": 0, "end": 3000}
  ],
  "language_code": "en"
}

Status values: queued → processing → completed (or error)

Configuration

All settings via environment variables with WHISPERX_ prefix. Everything has sensible defaults - no .env file needed for local use.

Speaker Diarization Setup

To enable speaker_labels=true:

1. Accept license at pyannote/speaker-diarization
2. Get token at huggingface.co/settings/tokens
3. Add to config:

   echo "WHISPERX_HF_TOKEN=hf_xxx" >> ~/.config/whisperx-api/.env
   

Troubleshooting

CUDA not available:

# Check NVIDIA driver
nvidia-smi

# Check PyTorch CUDA
python -c "import torch; print(torch.cuda.is_available())"

Out of VRAM:

• Use smaller model: WHISPERX_MODEL=medium
• Reduce batch size: WHISPERX_BATCH_SIZE=8

Diarization fails:

• Verify HF token: echo $WHISPERX_HF_TOKEN
• Accept license at HuggingFace (link above)

Built On

This project wraps the incredible WhisperX by @m-bain - fast automatic speech recognition with word-level timestamps and speaker diarization.

git clone https://github.com/namastexlabs/whisperx-api.git
cd whisperx-api
uv sync

uv run pytest tests/ -v

uv run ruff check .
uv run ruff format .
uv run mypy src/

whisperx-api/
├── src/whisperx_api/
│   ├── server.py          # FastAPI application
│   ├── transcriber.py     # WhisperX pipeline
│   ├── model_manager.py   # GPU model caching
│   ├── database.py        # SQLite persistence
│   ├── config.py          # Settings management
│   ├── auth.py            # API authentication
│   ├── models.py          # Pydantic schemas
│   ├── deps.py            # Dependency checks
│   └── main.py            # CLI entry point
├── tests/                 # Test suite
├── get-whisperx.sh        # One-liner installer
└── pyproject.toml         # Project config

Made with ❤️ by Namastex Labs