Sora 2 T2V

Sora 2 (text-to-video) is OpenAI’s model that converts natural language prompts into short, dynamic video clips with synchronized audio. You supply a descriptive prompt—such as “a cat chasing a butterfly across a garden at dusk”—and the model generates a video sequence that interprets scene composition, motion, lighting, and ambiance. The output includes matched audio layers: ambient sound, effects, and dialogue when relevant. The system aims for consistency across frames, realistic transitions, and alignment between the visual content and the narrative implied by the prompt.

October 6, 2025

OpenAI

Text to Video (T2V)