Ovi (with audio 5 seconds)

Ovi is an audio-video generation model developed by Character AI (in collaboration with Yale University) that produces short, cinematic video clips with built-in synchronized audio from text prompts Simultaneously creates realistic speech, sound effects, ambient audio and lip-synced visuals through its twin-backbone fusion architecture. Supports multi-person dialogues, audio descriptions tags (<AUDCAP>…<ENDAUDCAP>) and speech tags (<S>…<E>). Use cases for Ovi span from creative social media clips, narrative vignettes, short dialogues, to stylised advertising or promotional videos. Since the output includes audio and visuals, it gives you a more complete “mini-scene” experience in just a few seconds.