Open source model focused on high-accuracy, multilingual speech transcription.
/v1/models/cohere/cohere-transcribe-03-2026