OpenAI introduces Sora 2 Video Generation AI Model and launches a new TikTok-like social app

 

On Tuesday, OpenAI revealed the Sora 2 video generating model, which includes a number of new features and upgrades. The artificial intelligence (AI) model comes nearly a year after the first Sora model was launched. The new AI video model will be offered for free through a new social media app modeled like TikTok. Access to the iOS-only app requires an invitation. The program lets users to create AI films of themselves and share them with their friends. Users may also create comparable movies of others using the "Cameo" function.

Sora 2 AI Model and New Sora iOS App Released
In a blog post, the San Francisco-based AI business introduced and described the new AI model and app. The most significant increase is in native audio creation, which now rivals Google's Veo 3. OpenAI dubbed Sora 2 the "GPT-3.5 moment for video," claiming it can now manage complicated motions while keeping realistic physics in films.

Sora 2 is here. pic.twitter.com/hy95wDM5nB

— OpenAI (@OpenAI) September 30, 2025

In a demo film, a gymnast performed flips on a balancing beam with realistic motions, demonstrating the model's potential. OpenAI further stated that Sora 2 does not morph things or distort reality to carry out a command. This would result in less hallucinations in the produced videos.

The output's controllability has also been increased, and Sora 2 can now obey directions that span several shoots while keeping the world state constant, the firm stated. Notably, the video generation model is believed to be the best at producing realistic, cinematic, and anime-style content.

Currently, the Sora 2 model is available through the Sora for iOS application. It is an invite-only app available to people in Canada and the United States. The new app's UI is similar to TikTok, Instagram Reels, and YouTube Shorts. The main difference is that all of the videos in this app are created using Sora 2.

This is the Sora app, powered by Sora 2.

Inside the app, you can create, remix, and bring yourself or your friends into the scene through cameos—all within a customizable feed designed just for Sora videos.

See inside the Sora app👇 pic.twitter.com/GxzxdNZMYG

— OpenAI (@OpenAI) September 30, 2025

Notably, Sora 2 video creation is presently free for all users who receive an invitation. In addition, after a user has obtained an invite to the app, they can make movies on the Sora website.

Users may submit a video of themselves and use the AI model's strong character consistency to place the character in any other scenario. Users may then submit these videos for their followers to view, like, comment on, and share. They can also "cameo" them. While browsing through the videos, the user may hit the Cameo button to capture the character from that movie and create a new video with them and a text prompt. Essentially, it allows users to make deepfakes of their pals.

"We started experimenting with the "upload yourself" option on the Sora team some months ago, and we all had a great time with it. "It felt like a natural progression of communication—from text messages to emojis to voice notes to this," the business stated.

Notably, before a person can create a video of themselves using the app, they must authenticate their identity. So, a person cannot utilize a random film of a celebrity to create movies of them. However, the Cameo function removes some of the precautions.

Surprisingly, OpenAI is not the only company studying the vertical scrolling AI-generated video app concept. Last week, Meta introduced the Vibe stream to the Meta AI app, which allows users to view short movies created using in-house AI models.