Nvidia released Nemotron 3.5 Transcribe on Wednesday, a free open-weight model capable of transcribing audio in up to 40 languages. The company also launched version 2 of Granary, its open-source speech dataset. Nemotron 3.5 handles recordings up to an hour long and distinguishes between eight speakers while translating 31 languages to and from English. The model claims to outperform rivals on benchmarks such as Hugging Face’s Open ASR Leaderboard.
These tools target developers building applications for business meetings and medical transcription, two of AI’s most established footholds in the workplace. Nvidia states developers can use the model and the updated Granary dataset alongside tools like its NeMo Speech model to customise apps to recognise an industry’s jargon in each language. One drawback is that Transcribe works with recorded audio rather than live streams. The company promises the 30 billion parameter model can run on a single chip, ranging from the Jetson AGX Thor robotics controller to Blackwell datacenter chips. Nvidia’s latest voice entry arrives during a busy period of AI voice releases, following Mistral’s Voxtral 3, ElevenLabs’ Realtime, and Salesforce’s Launa-Omni.
* The model supports up to 40 languages
* Recordings can last up to one hour
* It runs on hardware from Jetson AGX Thor to Blackwell chips
