0
Applied AI·October 28, 2026·1 min read

Nvidia debuts Nemotron 3.5 Transcribe, an open-weight model for multilingual audio transcription and translation, and Granary v2, its open-source speech dataset

Share

High-quality, open-weight multilingual transcription plus an open dataset compresses the moat around proprietary speech stacks. If you’re paying per-minute for transcription at scale, it’s time to cost out a Nemotron-based path and ask vendors how they’ll justify their margin once customers can self-host.