IT之家 · ITHome
NVIDIA releases Nemotron 3.5 Transcribe with support for 40 languages
NVIDIA released the open-weight Nemotron 3.5 Transcribe model and made its Granary v2 training-data pipeline public. The model combines transcription, speaker attribution, translation and audio question answering. It handles recordings up to one hour, transcribes 40 languages, supports translation between English and 31 languages, distinguishes up to eight speakers and provides speaker labeling in 13 languages.

What to know
- 01
The model is freely available on Hugging Face and ModelScope under the OpenMDW-1.1 license.
- 02
Granary v2's published data inventory covers 463,062 hours of audio, 48 languages and 16 source corpora.
Why it matters
This release is relevant to transcription and speaker-labeling tasks in meetings, classroom recordings and multilingual customer service.
AI-assisted brief. Read the source for full details. Editorial & sourcing policy


