Screenshot by Radhika Rajkumar/ZDNET Follow ZDNET: Add us as a preferred source on Google. ZDNET's key takeaways OpenAI's most capable video model, Sora 2, is here. The company also launched a new iOS social media-like app. Both the app and the new model are free to access. If you thought OpenAI's first video-generating model, Sora, was realistic, wait until you see what Sora 2 can do on both the video and audio front. Also: Luma AI created an AI video model that 'reasons' - what it does differently OpenAI finally launched the highly anticipated next-generation flagship video and audio generation model, Sora 2, on Tuesday. The new model is meant to be significantly more capable, tackling typically difficult tasks for video generators, which OpenAI equates to the jump from GPT-1 to GPT-3.5. (Disclosure: Ziff Davis, ZDNET's parent company, filed an April 2025 lawsuit against OpenAI, alleging it infringed Ziff Davis copyrights in training and operating its AI systems.) Sora 2 does what Sora couldn't The new model was trained to tackle the challenges that plagued the previous model. For example, OpenAI said in a blog post that the model was trained to be less overly optimistic, a characteristic that can be observed in instances where a Sora-generated video shows the player missing the shot but still making it into the hoop. With Sora 2, OpenAI claims the player would miss the shot, and the ball would rebound off the backboard. The model is also designed to better adhere to the laws of physics and have greater controllability, allowing it to follow more complex instructions and achieve more realistic results overall. Also: I used Google's photo-to-video AI tool on my selfie - and it made me do the tango A major leap forward is its ability to create sound that pairs with the video. This includes sound effects, backgroundscapes, and even human speech. Google's Veo 2 video generator, unveiled in April, has the same video and audio capability, and it is stunningly realistic. Another nuanced feature is that users can now add real-world clips to the Sora 2 model. For example, OpenAI included a video of its teammates using AI to transform videos of themselves, demonstrating how they created a clip of a person chasing an ostrich or playing the trumpet in a sea of zebras. Compared to the other clips that OpenAI showed of Sora 2 in action, the ones building off of user video look a bit less realistic, but still fun nonetheless. Copyright opt-out Though not mentioned in OpenAI's announcement, The Wall Street Journal reported Monday that OpenAI will allow copyrighted material to be used in videos Sora 2 generates unless copyright holders opt out. According to WSJ, movie studios and other relevant parties that own intellectual property are required to ask OpenAI not to use their material. Also: Anthropic agrees to settle copyright infringement class action suit - what it means The requirement is the latest development in the ongoing legal battle between IP holders and AI companies that scrape text, images, and video off of the internet -- or, increasingly, license material from studios and publishers -- to train their models. New iOS social app In tandem with the new model, OpenAI also launched a new social iOS app called Sora and powered by Sora 2. Beyond creating and generating video in the app, there is a social media component that allows you to interact with others. Also: ChatGPT can buy stuff for you now - forever changing online shopping Much like any other video social app, with the Sora app, you can discover other people's creations in a customizable Sora feed, remix your friends' generations, and even drop yourself into your friends' creations via a cameo feature. Likeness protection OpenAI reassures users that they are in control of their likeness in cameos. Only you can decide who can use your cameo, revoke access, and remove yourself from videos created using your cameo. You can also view videos containing your cameo, even if it's in the other user's draft. For more on safety, you can refer to OpenAI's Sora 2 Safety doc⁠. Also: How to use ChatGPT freely without giving up your privacy - with one simple trick With a short one-time video and audio recording, the cameo feature can recreate your likeness in any Sora scene. OpenAI said the purpose of the one-time video and audio recording is to verify your identity in addition to capturing your likeness. Personalized algorithms To avoid the doomscrolling that people engage in as a result of their addiction to social media, the feed algorithms in the Sora app can be customized by the user through a new type of recommender algorithm that is instructed in natural language. Users will also be presented with check-ins asking about their well-being and will have the option to opt out of their feed. Without tweaking the algorithm, by default, the content you are shown will be mostly people you follow or interact with, and would also include recommended content that might serve as inspiration for your future projects. Also: Project Liberty's plan to decentralize TikTok could be the blueprint for a better internet "We are not optimizing for time spent in feed, and we explicitly designed the app to maximize creation, not consumption," said OpenAI. The company also carefully crafted the teen experience in the Sora app, including limits on the number of generations they can see per day in their feed, stricter permissions on cameos, and larger teams of human moderators to identify bullying. OpenAI has also launched Sora parental controls via ChatGPT, allowing parents to customize their children's experience, including overriding infinite scroll limits, disabling algorithm personalization, and managing direct message settings, according to the blog. Pointing to the problem of social media apps monetizing attention, OpenAI said its only plan to make money off the Sora app is to have users pay an undisclosed amount extra -- the app is free to use to start -- to generate more video if there's too much demand relative to the available compute. Also: I teamed up two AI tools to solve a major bug - but they couldn't do it without me The app is rolling out on an invite-only basis, a decision OpenAI said was made to ensure users join with friends. How to access Sora 2 (and the new app) The Sora iOS app is now available for download from the Apple App Store. You can then sign up for push notifications when access to your account opens. The app is rolling out to the US and Canada today, but OpenAI said it will expand to additional countries. Once you have received an invite, you'll be able to access Sora 2 via the Sora website. The Sora 2 experience will be free to begin with, but ChatGPT Pro users will also be able to use the higher-quality experimental Sora 2 Pro offering on the Sora website and soon on the Sora app.