Our latest video generation model is more physically accurate, realistic, and more controllable than prior systems. It also features synchronized dialogue and sound effects. Create with it in the new Sora app.
OpenAI has released Sora 2, a new video + audio generation model that improves realism, physical consistency, and controllability. The system better preserves world state across shots, obeys physics (for example a ball rebounds instead of teleporting), and supports synchronized speech plus sound effects. OpenAI also launched a companion iOS app for social video creation, remixing, and “cameos” that let you insert yourself into scenes. They built safety controls, parental limits, and consent tools into the app. Initially Sora 2 is rolling out in the U.S. and Canada, free with limits, and will expand globally with plans for a pro version and API access.