AIThe New Stackabout 1 hour ago
Nvidia Launches Promptable Transcription Model
3 min read
Nvidia launches Nemotron 3.5 Transcribe, a 30B model with 4 billion active parameters, on Wednesday. It ranks first on Hugging Face's Open ASR leaderboard with a 5.33% word error rate. The system handles 40 languages and distinguishes up to eight speakers. Nvidia also releases Granary v2, a pipeline covering 844,000 hours of speech across 48 languages.
Level
Hype check
Tap to vote and see what everyone thinks.
In this storyNvidia
