← Back to feed

Aura: 1.3B speech-to-video model that runs in real time on a single H100 GPU

AI · Technology · 2026-11-20 15:35

Kyutai has released AURA, a 1.3-billion-parameter speech-to-video model capable of generating video in real time from audio input on a single H100 GPU.

Source: huggingface.co


Auto-generated summary for reference only. RSS