Our video model Sora is capable of creating videos from text, images, and other videos, but is also capable of changing just one element of a video, like these examples where just one variable is changed in each of these generations. These videos were generated by our AI video model, Sora, without modification. What would you like to see us make with Sora next? *Sora is not yet available to the public. We’re sharing our research progress early to learn from feedback and give the public a sense o

AI Has HIT ITS STRANGEST MOMENT YET 💀 (Sora 2 Clips That Make No Sense)

AI Has ENTERED FULL BREAKDOWN 💀 (Sora 2 Moments That Didn’t Recover)

AI Has FINALLY GONE OFF THE EDGE 💀 (Sora 2 Moments You Never Expected)