OpenAI has released its latest video generation model, Sora 2. This new model is more physically accurate and controllable than prior systems, with synchronized dialogue and sound effects. It can create realistic backgrounds, speech, and sound effects, and allows users to inject real-world elements into generated environments.
Sora 2 represents a significant leap forward in video generation, capable of creating sophisticated content that’s comparable to human-made videos. The model excels at realistic, cinematic, and anime styles, making it a valuable tool for creators and artists.
To ensure user wellbeing, OpenAI has implemented safety features such as customizable feed options, strict permissions on cameos (user-controlled likenesses), and automated moderation teams to review cases of bullying. The app also prioritizes creation over consumption and provides users with control over their content.
Sora 2 is now available for download on the iOS app store, initially in the US and Canada, with plans to expand globally. The model will be free, with generous limits, and users can access it through sora.com or the Sora app. OpenAI also plans to release Sora 2 as an API for developers.
The release of Sora 2 marks a significant step towards general-purpose world simulators and robotic agents that can fundamentally reshape society. With its groundbreaking capabilities, this model has the potential to bring joy, creativity, and connection to people worldwide.
Source: https://openai.com/index/sora-2