Voice AI in vogue: Nvidia releases new open transcription model

Nvidia just released a free speech-to-text model that can transcribe up to 40 languages, but it only works on recordings, not live audio, so it’s a big step for meeting and...