Cohere Transcribe is an open source research release of a 2B parameters dedicated audio-in, text-out, automatic speech recognition (ASR) model. The model supports a total of 14 languages.
Not enough throughput data
Not enough TTFT data
