Whisper Large v3 Turbo

Permissive

OpenAI · Whisper · Released 2024-10

809MAudio

A pruned-decoder version of Whisper large-v3 that runs several times faster with only a small accuracy tradeoff, good for real-time use.

Strengths

  • +MIT license, unrestricted use
  • +Significantly faster inference than large-v3
  • +Good accuracy retention despite the speedup

Limitations

  • -Slightly lower transcription accuracy than full large-v3
  • -Translation quality (non-English to English) trails the full model

License

MIT

use commercially with attribution niceties

Hardware

runs on a good consumer GPU (or Apple Silicon) with quantization

Links

Stats

openai/whisper-large-v3-turbo— downloads·— likes

via Hugging Face

Related models