OpenAI’s TikTok Rival Sora Adds Social Feed Pet Cameos and Android Support

OpenAI’s viral AI video app Sora is expanding beyond human cameos. The company announced on Tuesday a series of updates that will transform users’ pets, stuffed toys, and other objects into AI-generated video characters. The enhancements arrive as Sora maintains its position at the top of the App Store in the U.S. and Canada, having reached one million downloads in under five days.

Bill Peebles, head of Sora at OpenAI, revealed the roadmap on X, promising that the new creation tools will roll out within days. The announcement marks OpenAI’s clearest signal yet that it views AI video generation as a social media battleground, directly challenging TikTok’s dominance in short-form content.

Expanding the Cameo Universe

Sora’s defining feature has been its “cameo” system, which allows users to upload reference videos of themselves and create AI personas that friends can use in their own generations. The upcoming update extends this concept dramatically. Users will soon be able to create cameos of dogs, guinea pigs, favorite stuffed toys, and virtually any object.

The expansion addresses a fundamental desire among early adopters. Rather than limiting AI character creation to human subjects, OpenAI is betting that users want to star their pets and possessions in fantastical scenarios. The update also introduces the ability to create cameos directly from AI-generated characters that appear in Sora videos, enabling consistent fictional personas across multiple generations.

To manage what OpenAI expects will be an explosion of new cameos, the company is redesigning the generation interface to display trending cameos in real time. This discovery mechanism mirrors the algorithmic recommendations that power TikTok’s For You page, suggesting OpenAI views viral content distribution as central to Sora’s growth.

Building a Social Ecosystem

Beyond creation tools, Sora is evolving its social infrastructure. Peebles emphasized that OpenAI is “working on making the social experience much better,” exploring ways for users to engage with Sora beyond the current global feed. The company is developing dedicated channels specific to universities, companies, sports clubs, and other communities.

The timing is strategic. Meta launched Vibes in late September, its own AI video feed within the Meta AI app, and has seen significant increases in downloads and daily active users. The competitive pressure from Meta, combined with Google’s Veo 3 and other emerging players, is accelerating OpenAI’s push to establish Sora as the dominant AI video platform.

Video Editing and Android Timeline

For the first time, Sora will include native video editing capabilities. The initial release focuses on the basic functionality of stitching together multiple clips within the app. OpenAI promises more powerful editing features in subsequent updates. This addresses a workflow friction point for users who currently must export Sora generations to third-party editors.

Since its September 30 launch, Sora has been exclusively available on iOS. Peebles confirmed that an Android version is “actually coming soon,” though he provided no specific launch date. The app has been available for pre-registration on the Google Play Store. Android support will be crucial for international expansion, particularly in markets where iOS penetration remains limited.

Addressing Moderation Backlash

The updates arrive as OpenAI grapples with persistent complaints about over-aggressive content moderation. Users have reported excessive blocking of legitimate creative content, with some finding that even public domain characters trigger violation warnings. Peebles acknowledged the company is working to “reduce the excessive moderation of generations” and improve overall app performance.

This represents a delicate balancing act. OpenAI launched Sora with relatively permissive policies that allowed users to generate copyrighted characters, then faced immediate backlash from the Motion Picture Association and individual rights holders. The company reversed course within days, shifting to an opt-in model for copyrighted content.

Viral Launch and Compute Crisis

Sora’s explosive debut has fundamentally altered OpenAI’s trajectory. The app reached one million downloads faster than ChatGPT, despite being invite-only and limited to North America. By its second day, Sora had claimed the number one position on the U.S. App Store, surpassing both ChatGPT and Google Gemini. Third-party estimates from Appfigures suggest the app has accumulated approximately two million downloads since launch.

This adoption reveals pent-up demand for consumer AI video tools packaged in social media formats. However, the rapid adoption creates an existential challenge for OpenAI. CEO Sam Altman acknowledged that “people are generating much more than we expected per user,” consuming massive computational resources for content often shared with very small audiences.

“We are going to have to make money for video generation somehow,”

Altman wrote in an early October blog post.

The company currently offers Sora for free to all users, with more advanced features reserved for paying subscribers. This unsustainable economics forced Altman to float potential solutions, including revenue sharing with rights holders who opt into allowing their characters in Sora videos. The compute burden of video generation dwarfs text or image generation, creating ongoing infrastructure challenges as the user base expands.

OpenAI’s rapid feature rollout signals its ambition to transform Sora from a creative tool into a full-fledged social platform competing directly with TikTok, Instagram Reels, and Meta’s Vibes but with entirely AI-generated content. Whether users will embrace a purely synthetic feed remains uncertain.

As OpenAI races to ship features, the company faces mounting pressure. Content moderation must satisfy both creators and rights holders, monetization must offset crushing compute costs, and Android support must arrive before competitors capture that market. Most fundamentally, OpenAI must prove that AI-generated social content represents more than a passing curiosity.