NVIDIA releases Nemotron 3.5 Transcribe and Granary v2
NVIDIA released the open-weight, promptable Nemotron 3.5 Transcribe model for recorded-audio transcription, translation, speaker attribution, and audio question answering, alongside the Granary v2 data-curation pipeline and fine-tuning resources. Reports say it supports 40 languages, translates 31 with English, labels up to eight speakers, and is available on Hugging Face and ModelScope, but it does not support live streaming. NVIDIA and its technical partners report strong benchmark results, while published sources differ on parameter counts, audio limits, and some benchmark figures.
The release gives developers an open-weight option that combines several recorded-audio tasks in one adaptable model and can be fine-tuned with Granary v2 for specialized terminology and transcript formats.
Nvidia Corp. today released a leaderboard-winning open artificial intelligence model for multilingual audio transcription and translation, along with an open data pipeline for fine-tuning Nemotron 3.5 Transcribe.
Traditional transcription AI models follow a workflow that detects speech, transcribes it, identifies and labels different speakers, then answers questions about the transcription. Transcribe does all of this in a single model pass and is capable of digesting and answering questions about 100 hours of audio, in 40 languages and can label up to eight speakers.
It’s an open-weight, 30-billion parameter mixture-of-experts model with 4 billion active parameters. That means it can fit on reasonably small hardware, though more powerful hardware lets it transcribe faster.
For its size, Transcribe punches way above its weight and it can run anywhere from the edge to the data center depending on compression. The model runs on a single Nvidia graphics processing unit, from a Jetson AGX Thor, an edge computer robot brain, or DGX Spark, a desktop AI supercomputer, all the way to a Blackwell data-center GPU. Nvidia said a Blackwell Server Edition GPU with NVFP4 precision and one-quarter model compression processes 100 hours of audio in under three minutes.
#NVIDIA#Nemotron 3.5 Transcribe#Granary v2#NeMo Speech#NeMo Curator