Sora is here
OpenAI ● Covered by 3 sources
OpenAI just flipped the switch on Sora at sora.com, letting anyone make AI videos. Up to 1080p and 20 seconds, in whatever aspect ratio you want — that's a real leap from text-to-video demos.
Based on reporting by OpenAI — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
OpenAI has stopped teasing and started shipping. Sora, the video generation model that spent over a year living in cinematic trailer clips and cherry-picked demo reels, is now sitting behind a plain URL anyone can visit: sora.com. No waitlist theater, no research-preview fine print stopping you at the door. You show up, you type a prompt, you get a video.
The specs are the part that actually matters here, because specs are what separate a toy from a tool. Sora now outputs up to 1080p resolution, clips as long as 20 seconds, and in three aspect ratios — widescreen, vertical, and square. That trio isn't random. Vertical is for TikTok and Reels, square is for feed posts, widescreen is for anything meant to look like it belongs on a real screen. OpenAI built this for how video actually gets consumed in 2024, not how it gets consumed in a keynote.
What's more interesting than the raw generation is the editing layer sitting on top of it. Sora isn't just a black box that spits out a finished clip from a caption. You can upload your own footage or images and use the model to extend a scene, remix it, or blend it with something else entirely. That's a meaningfully different pitch than
My take — AI-written commentary, not fact-checked reporting
type words, get video.
Read more about this at: OpenAI