NVIDIA Releases Nemotron 3.5 Transcribe for Recorded Speech
Sign in✓ Signed in
NVIDIA has released Nemotron 3.5 Transcribe, an open-weight model for transcribing recorded audio in up to 40 languages. It also supports translation and speaker-attributed transcripts, but does not accept live audio streams. NVIDIA's model documentation lists recordings up to one hour, word timestamps, language identification, and audio question answering among its capabilities.
Continue reading
The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.
Already purchased? Sign in✓ Signed in — this article isn’t included in your current plan.
Unlocking the full article...