Sora is OpenAI's AI-powered text-to-video and image-to-video generation tool, designed to turn written prompts and still images into high-definition videos with hyperreal motion and synchronized audio. Powered by the Sora 2 model, the platform enables anyone to create cinematic scenes, animated shorts, photorealistic clips, or entirely surreal content — all from a single sentence or uploaded image.

Users can generate videos up to 20 seconds long at up to 1080p resolution in widescreen, vertical, or square aspect ratios. The model produces synchronized music, sound effects, and dialogue automatically, making every generated scene complete without additional editing. Sora builds on years of research into diffusion transformers and large-scale video data training, resulting in a system that accurately simulates real-world physics, consistent object permanence, and natural motion across complex scenes.

The Sora app includes social and collaborative features: users can cast themselves or friends as characters in any video through a one-time likeness capture, remix existing creations by swapping characters or changing the vibe, and discover a personalized community feed of AI-generated videos. Storyboard tools allow frame-by-frame control over complex video sequences, while the stitching feature lets users connect multiple clips into a longer story. All generated videos include a visible moving watermark and embedded C2PA metadata to identify AI-generated content. Sora is accessible on the web at sora.com, on iOS, and on Android.