OpenAI has rolled out its latest AI innovation, Sora 2, a video generation model that's already turning heads in the tech world. Launched just yesterday on October 1, 2025, the new Sora app promises to make short video creation as easy as scrolling through your feed. But here's the catch: it's invite-only, and those coveted codes are flying around online forums like digital gold.
The app, available now on iOS, lets users whip up 10-second vertical clips from simple text prompts or photos, complete with synchronized audio and sound effects. OpenAI touts improvements in realism and physics simulation, making videos feel more lifelike than ever. Indeed, it's powered by Sora 2, which builds on the original model's strengths but adds better control over elements like movement and dialogue. For ChatGPT Pro subscribers, there's even an experimental higher-quality version rolling out soon on the web and app.
However, access remains tightly controlled. During onboarding, users must enter an invite code to bypass the waitlist—though OpenAI says you can join that list directly in the app if you're shut out. Social media is buzzing with exchanges; Reddit threads alone have thousands commenting, begging for spares or offering trades. One post I saw had over 14,000 replies, a testament to the hype. And while it's fun to imagine friends remixing AI videos of each other in a TikTok-style feed, the exclusivity rubs some the wrong way. Why gatekeep creativity like this when AI is supposed to democratize tools?
Moreover, the app emphasizes collaboration: follow friends, share clips, and build on ideas together. It's a bold pivot for OpenAI, blending video gen with social networking in a way that could rival bigger players. But for now, Android users are left waiting, as the native app is iOS-exclusive at launch. Codes seem to work on the web version too, offering a workaround.
In this rush to innovate, OpenAI's Sora 2 app highlights both the thrill of AI's creative potential and the barriers still in place for most users.