OpenAI's Sora 2 can generate cinematic videos with immersive sound
OpenAI just rolled out Sora 2, its newest text-to-video model, on September 30, 2025. It's a big leap from the last version and goes head-to-head with Google's Veo-3.
Sora 2 can handle complex multi-scene prompts and keeps everything looking consistent—think realistic, cinematic, or even anime vibes, all with immersive sound effects.
There's also a new Sora app for iOS
Alongside the model, OpenAI launched the Sora app for iOS.
The app lets you make and explore personalized video feeds—and with its "upload yourself" feature, you can actually put yourself into scenes by recording a quick video or audio.
Content recommendations are powered by natural language AI that highlights creators you follow and fresh inspiration, making it easy to discover cool new videos while keeping things in your control.