WhisperX
Fast transcription with word-level timestamps and speaker diarization.
- Star 数
- 24.6K
- Fork 数
- 2.5K
- 未关闭 Issue
- 230
- 最近提交
- 2026-09-26
数据来自 GitHub API · 更新于 2026-10-15
WhisperX 是什么？
WhisperX combines faster-whisper batched inference with forced alignment for accurate word timestamps and pyannote speaker diarization — ideal for meeting transcripts and subtitles.
主要功能
- Batched inference for speed
- Accurate word-level timestamps via alignment
- Speaker diarization
系统要求
- 平台
- Windows · macOS · Linux
- 显卡
- NVIDIA recommended
- Python
- 3.9+
- 备注
- Diarization needs a Hugging Face token with access to pyannote models.
如何安装 WhisperX
根据项目 README 整理。开源项目更新很快，请以 GitHub 页面上的最新说明为准。
pip
- 1Install.
pip install whisperx - 2Transcribe with speaker labels.
whisperx audio.wav --model large-v2 --diarize --hf_token YOUR_HF_TOKEN
优势
- Best timestamps and speaker labels
- Fast
不足
- More dependencies to set up
WhisperX 替代品对比
| 软件 | Star 数 | 许可证 | 平台 | 显存 |
|---|---|---|---|---|
| WhisperX Fast transcription with word-level timestamps and speaker diarization. | 24.6K | BSD-2-Clause | — | |
| faster-whisper Whisper reimplemented on CTranslate2 — up to 4× faster with less memory. | 25.9K | MIT | — | |
| Whisper OpenAI’s open-source speech recognition and translation model. | 110.5K | MIT | ~1 GB (tiny) to ~10 GB (large) | |
| whisper.cpp Whisper in plain C/C++ — fast on CPUs, Apple Silicon and mobile. | 54.4K | MIT | — | |
| Coqui TTS Deep-learning text-to-speech toolkit, including the XTTS voice cloning model. | 46.1K | MPL-2.0 | — | |
| GPT-SoVITS Few-shot voice cloning and TTS with a full training WebUI. | 62.8K | MIT | 6 GB+ |
常见问题
WhisperX 免费吗？
是的——WhisperX 是开源软件，采用 BSD-2-Clause 许可证。商用前请确认软件及模型权重的许可条款。
WhisperX 支持 Windows、Mac 和 Linux 吗？
WhisperX 支持：Windows, macOS, Linux。
运行 WhisperX 需要什么配置？
NVIDIA recommended
如何安装 WhisperX？
1. Install. 2. Transcribe with speaker labels.
WhisperX 有哪些好的替代品？
常见替代品包括 faster-whisper, Whisper, whisper.cpp, Coqui TTS，详见本页对比表。