Openai today announced the launch of Sora, an invitation-only AI video app and social network. SORA allows you to create realistic AI videos for yourself, your friends and others.
SORA uses the SORA 2 video generation model. Openai says it is physically more physically accurate, realistic and controllable than its previous systems. It can produce complex movements while better following the laws of physics, Openai says it is excellent in its realistic, cinematic and anime style.
SORA 2 supports the generation of videos that also include audio, including actual audio, background soundscapes, and sound effects.
AI models can accurately depict the appearance and voice that is the basis of new SORA apps, while observing people's videos and inserting them into the SORA generation environment.
The SORA app allows you to create your own videos that can be inserted into “Cameos,” a short video that will be shared with others on the Sora social network. You can choose to allow others to create cameos with your portrait. You can choose who can use the cameo. You can also watch all videos with your portraits, including cameos, drafts before they are published to the network.
Openai designs SORAs, displays content based on who you follow or interact with, and the app regularly votes about your happiness. There is a control that changes what is shown in the feed, and Openai says it is intended to be used with friends. Sora invites only invitations and makes sure people join the app with people they know.
The SORA app for iOS is available for download now and can be used in the US and Canada. Those invited to the app can use Sora 2 with SORA 2.
The SORA 2 is free for now, and ChATGPT Pro users can access the SORA 2 Pro model on SORA.com.
