OpenAI has released Sora 2, a video creation model that takes AI video to new heights. For now, access is limited, the app and website are free, with usage limits based on available processing power.
Sora 2 addresses the issues with previous AI video models by focusing on physical realism. Unlike the first Sora, which was mostly visual, this one generates synchronized conversation, sound effects and background soundscapes. A prompt for two mountain climbers shouting in a snowstorm yields pictures of wind whipped coats, ice crusted faces and the crunch of snow as well as desperation in their voices. You can control the model with specific commands that span multiple shots while having consistent characters, settings and actions, across scenes.
- Faster Performance for Everyday Computing: Powered by the Intel Celeron N5095 processor (4 Cores, 4 Threads, up to 2.9GHz), the KAMRUI AK1PLUS Mini PC...
- 16GB RAM & Up to 4TB Expandable Storage: Work faster with 16GB LPDDR4X memory and a high-speed 256GB M.2 2280 SSD for quick boot-ups and smooth...
- Dual 4K UHD Displays for Maximum Productivity: Expand your workspace with dual HDMI 2.0 outputs, supporting two 4K@60Hz UHD displays simultaneously...
The new Sora iOS app, launched with the model, adds a social dimension to video creation. It’s an invite only site that will first be available on iOS in the US and Canada. You can make 10 second videos, browse a curated feed and remix other people’s work. The stream shows content from people you follow or videos that will inspire your own thoughts so you don’t get lost in endless scrolling.
Another cool feature is “Cameos” which allows you to put yourself or others into videos. After a single video and audio recording to confirm your identity, the model will transport you to any setting, whether you’re hiking a snowy mountain or befriending a friendly Bigfoot. OpenAI prioritizes user control: you decide who can use your image and you can cancel or delete any video of you at any time.
Physics with Sora 2
…and some anime. pic.twitter.com/AkqVysryzH
— OpenAI (@OpenAI) September 30, 2025
ChatGPT Pro users get the higher quality “Sora 2 Pro” model, while Sora 1 Turbo is available to everyone else. OpenAI is working on an Android app and will roll out to more countries soon. The API will likely allow other developers to experiment with the model.
[Source]
