Background
Member Login
menu

OpenAI Gives Sora 2 'Character Cameos' Feature, And The Ability To Stitch Videos Together

Sora Character Cameos

The LLM war isn’t just about chatbots any more. What once started as model vs model in text has now spilled into video generation.

What began from the release of ChatGPT form OpenAI, others followed suit. Now, Google is pushing models such as Veo 3 and Veo 3.1, which can generate videos with synchronized audio to stake their claim in the new creative frontier. Meanwhile, OpenAI that started its video-generation trend with Sora, has introduced Sora 2.

OpenAI positions Sora 2 not just as a tool, but as a social video platform where creation, remixing and sharing come together.

While Google’s Veo emphasizes high-fidelity audio-video sync and traditional production workflows, Sora 2 takes the virality route: short-form, remix friendly, avatar-ready, social feed enabled.

And now, OpenAI has updated Sora 2 to include its biggest social-creative lever yet: 'Character Cameos.'

This feature allows users to turn almost anything, like a pet, a drawing, a toy, even an illustration, into a reusable avatar (a "cameo") that can be tagged in future AI-generated videos.

According to OpenAI in the Sora release notes on October 29:

"Once created, each character comes with its own permissions, separate from your personal likeness: keep it just for yourself, share it with mutual followers, or open it to everyone on Sora. Give your character a display name and handle, and tag it whenever you want it to appear in a video."

Character Cameos were first hinted at a week earlier, and OpenAI explained that the feature builds upon Sora’s existing system that lets users create AI-generated versions of themselves, digital doubles that others can use, provided they’ve given permission.

Now, those same deepfake-style capabilities extend beyond people to include virtually anything: pets, drawings, toys, or other imaginative subjects. To kick things off, OpenAI is also rolling out a starter pack of pre-made characters that anyone can instantly drop into their videos.

And this makes a lot of difference in the playing field.

To get users going, OpenAI is releasing the feature with starter pack characters for users to play with, including classic Halloween character cameos like:

  • Frankenstein (@franklyfrankenstein).
  • Dracula (@soradracula).
  • Jack O Lantern (@sorajackolantern).
  • Witch (@sorawitch).
  • Ghost (@ghostlyghost).

Beyond cameos, Sora 2 also introduces video stitching.

In both the Sora app and at sora.com, users can select multiple short clips to be joined into longer, multi-scene sequences, and leaderboards to surface the most remixed videos and most used characters.

These features together are designed to turn creation into a collaborative, dynamic affair. They allow not just one-off videos to be created and shared, but evolving characters and even threads.

"In drafts, tap ‘select’ and choose the clips you want to stitch together. Then, click ‘stitch’ to connect the videos. The stitched video will appear in your drafts once ready. Stitching allows you to create more sophisticated narratives and extend your story, with granular control over each segment," explained OpenAI.

But this new versatility comes with controversy.

The "cameo" feature is already at the center of a legal dispute, where the celebrity video platform Cameo has filed a trademark infringement lawsuit against OpenAI, claiming the name "Cameo" in Sora’s context creates confusion with its own brand of celebrity video messages.

On another front, Sora 2 has been criticized for its deep-fake potential, because videos of well-known figures (including deceased ones) are emerging, raising serious ethical and consent questions.

Published: 
01/11/2025