OpenAI today announced the launch of Sora, an invite-only AI video app and social network. Sora lets you create realistic AI videos of yourself, friends, and other people.
Sora uses the Sora 2 video generation model, which OpenAI says is more physically accurate, realistic, and controllable than prior systems. It is able to generate complex movements while better obeying the laws of physics, and OpenAI says it excels at realistic, cinematic, and anime styles.
Sora 2 supports generating video that also includes audio, such as real-sounding speech, background soundscapes, and sound effects.
The AI model is able to observe a video of a person and then insert them into a Sora-generated environment while accurately portraying their appearance and voice, which is the basis for the new Sora app.
With the Sora app, you can create a video of yourself that can then be inserted into "cameos," which are short videos that are shared with others on the Sora social network. You can opt to allow other people to create cameos with your likeness as well. You can choose who can use your cameo, and you will see all videos that include cameos with your likeness, even drafts before they are published to the network.
OpenAI designed Sora to show you content based on people you follow or interact with, and the app will poll you regularly on your wellbeing. There are controls to modify what's displayed in a feed, and OpenAI says that it is meant to be used with friends. For that reason, Sora is invite only, ensuring people join the app alongside people they know.
The Sora app for iOS is available to download now, and it can be used in the United States and Canada. Those invited to the app will be able to use Sora 2 on the Sora website.
Sora 2 is free for now, and ChatGPT Pro users have access to the Sora 2 Pro model on Sora.com.
