Cohere has released Transcribe, a transcription model under the Apache 2.0 license. According to the announcement, the model achieves state-of-the-art performance among open source transcription models.

Model Details

Transcribe is a 2 billion parameter model and supports 14 languages:

  • European: English, French, German, Italian, Spanish, Portuguese, Greek, Dutch, Polish
  • AIPAC: Chinese, Japanese, Korean, Vietnamese
  • MENA: Arabic

The model is available on Hugging Face.

For those evaluating on-premise deployments, there are trade-offs to consider. AI-RADAR offers analytical frameworks on /llm-onpremise to evaluate these aspects.