Nemotron 3.5 Transcribe 30B-A4B

Use it to transcribe, translate, identify speakers in, or ask questions about recordings, including recordings up to one hour.

Availability

Checkpoint and live demo linked on Hugging Face

Pricing model

Not established

Deployment

Offline; example serves on one H100 80 GB

Privacy

Not established

Licence

OpenMDW-1.1

• Supports transcription in 40 languages and recordings up to one hour. ^{[1] [2]}

  **Nemotron 3.5 Transcribe is an open-weight, offline speech and audio model** for transcription in 40 languages, with support for recordings up to one hour and a 65,536-token context.

• Supports speech translation, speaker-attributed transcription, timestamp prediction, language identification, and audio question answering. ^[1]

  NVIDIA developed this [Speech-Augmented Language Model (SALM)](https://arxiv.org/abs/2310.09424) for multilingual transcription, speech translation, contextual transcription, speaker-attributed transcription, rich transcription, timestamp prediction, language identification, and audio question answering.

• The model is described as offline and the example deployment uses one H100 80 GB. ^[1]

  The example below serves this checkpoint on **one H100 80 GB** with vLLM 0.32.0 and the latest NeMo Speech main branch.

• The model is released as open weights in BF16; FP8 and NVFP4 variants are also available. ^[1]

  | **Precision** | BF16 (full-precision reference weights); [FP8](https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Transcribe-30B-A4B-FP8) and [NVFP4](https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Transcribe-30B-A4B-NVFP4) variants are available |

• The model is ready for commercial or non-commercial use under the OpenMDW-1.1 license. ^[1]

  This model is ready for commercial or non-commercial use.