← Back to feed Brand & Business

Cohere launches an open source voice model specifically for transcription

TechCrunch 26 March 2026 12h ago
Cohere launches an open source voice model specifically for transcription
64
Relevance
9/25
Freshness
25/25
Authority
25/20
Brand Signal
3/15
Depth
2/15
Relevance Freshness Authority Brand Depth
Relatively light at just 2 billion parameters, the model is meant for use with consumer-grade GPUs for those who want to self-host it. It currently supports 14 languages.
Read Full Article → TechCrunch ↗