TLDR Tech / IT
Cohere launches open-source speech-to-text model for enterprise transcription and real-time workflows
Cohere has released an open-source, 2B parameter speech-to-text model designed for enterprise environments. The model prioritizes real-world accuracy and low latency, currently leading Hugging Face leaderboards. It aims to improve transcription in noisy, multi-speaker conditions, with future plans to integrate directly into Cohere's North enterprise AI platform.
Read the full story