OpenAI Turbocharges Sora AI Video Generation Tool And Makes It Public

OpenAI continues to roll out new products and pricing tiers as part of its 12 Days of OpenAI event, with the company releasing Sora, its video generation model, to general users. Prior to this release, Sora was available to a limited number of users as part of a research preview. The company says that this new model “serves as a foundation for AI that understands and simulates reality—an important step towards developing models that can interact with the physical world.”

Sora is a model that is able to “take text, image, and video inputs and generate a new video as an output.” Users will even have the ability to “enhance, remix, and blend their own assets.” Although the model is limited in the length and resolution of the video it’s capable of generating—it can only kick out videos up to 20 seconds long, while the resolution is capped to 1080p.


There will be two different versions of Sora for users to choose from: Sora and Sora Turbo. The latter model will be for those who are subscribed to the company’s ChatGPT Plus and Pro tiers. Those on the Plus plan will be limited to generating 50 videos at 480p, and less at 720p. Meanwhile, Pro plan subscribers will receive 10 times as much storage and will be able to generate higher resolution videos with longer lengths. Interested users can visit sora.com to begin to use these models.

The company points out that this version of Sora “has many limitations.” The model “often generates unrealistic physics and struggles with complex actions over long durations.” Therefore, users need to go in with realistic expectation as to what’s possible with this new model.

While it’s exciting to see OpenAI release more products, it feels like this one might be a tad too early. With the limitation on length, it’s tough to envision how anything other than some concept videos are possible. Hopefully the community can come up with something interesting to justify the release of Sora.