Google Veo 3

Google DeepMind's state-of-the-art text-to-video model (Veo 3.1): generates high-fidelity video with natively synchronized dialogue and sound effects.

Veo 3.1 is Google DeepMind's premier generative AI model for high-fidelity video creation. It delivers 1080p video with best-in-class realism, physics, and prompt adherence. The critical update: native audio generation, which includes synchronized dialogue, sound effects, and ambient noise. Available via the Gemini API and Google Vids, Veo 3.1 provides advanced creative controls (reference images, video extension) for filmmakers and content teams to produce professional-grade content efficiently.

https://ai.google.dev/models/veo-3-1