Nvidia Nemotron 3.5 Transcribe: open ASR for 40 languages
NVIDIA has released Nemotron 3.5 Transcribe, an open-weight model capable of multilingual transcription, speaker attribution, and audio Q&A across 40 languages. The release includes the Granary v2 data pipeline to help developers adapt the model for specific domains.
Why it matters
This release provides developers with powerful, open-weight tools for advanced audio processing, potentially lowering the barrier for high-quality multilingual transcription services.
Back to Articles a]:hidden"> NVIDIA Nemotron 3.5 Transcribe and Granary v2: Intelligent Multilingual Transcription You Can Prompt and Adapt Enterprise + Article Published October 28, 2026 Upvote 14 +8 Maryam Motamedi maryameee Follow nvidia Piotr Żelasko piotrzelasko Follow nvidia Kunal Dhawan kunaldhawan Follow nvidia Nikolay Karpov nkarpov Follow nvidia Nune Tadevosyan nunetadevosyan Follow nvidia Meline Mkrtchyan mmkrtchyan Follow nvidia Nithin Rao Koluguri nithinraok Follow nvidia Francesco fciannella Follow nvidia :last-child]:mb-0"> One open model for multilingual transcription, speaker attribution, translation, and audio Q&A, with top leaderboard results, plus the open Granary v2 data pipeline and fine-tuning recipes to adapt it. Domain gains, measured on held-out speakers One open model for multilingual transcription, speaker attribution, translation, and audio Q&A, with top leaderboard results, plus the open Granary v2 data pipeline and fine-tuning recipes to adapt it.
Also covering this story
One other newsroom covered this event. We read that version too.
Get smarter about the news
Sign up free for a feed built around what you actually care about, Dive Deeper research on any story, and the full text of every article.
Create free accountAlready have an account? Sign in