Nvidia releases free 100M-parameter model that identifies up to 8 speakers in real time

The Decoder · rss · 2026-09-27

Nvidia has released Nemotron 3 Diarization, a free AI model that performs speaker diarization—identifying who is speaking at any given moment in a conversation.

The model is only about 100M parameters yet can distinguish up to eight speakers in real time, making it a lightweight, practical building block for transcription, captioning, and voice-analytics applications.

Original post →

More from Models

Models channel →