Hello Creators,
OpenAI just unveiled its boldest move into the social space yet: the Sora app — a generative video platform powered by the brand-new Sora 2 model. Think TikTok, but built entirely around AI-generated video, with a focus on personalization, collaboration, and safety.
What’s inside the Sora app?
- Generative video feed — create and share short, dynamic clips of yourself or others
- Cameos — record a sample of your face and voice, then grant permission for others to feature your likeness in their videos
- Consent controls — you can revoke cameo access at any time
- Parental settings — limit personalization and messaging for younger users
It’s designed to feel familiar (a scrolling feed of creative clips) but built on top of generative video models instead of traditional uploads.
Sora 2: The upgraded model behind it
The app runs on Sora 2, OpenAI’s next-gen video model with big improvements in realism and stability.
- Physics-aware motion — objects move naturally without awkward warps
- Stable, believable scenes — from gymnastics to diving, movements look fluid
- Better consistency — subjects hold their shape across frames, reducing visual glitches
Access and rollout
- App — invite-only for now, launching first in the U.S. and Canada
- Model — available immediately to ChatGPT Pro users (no invite required)
- Free at launch — though OpenAI has hinted at monetization tied to video generation during peak demand
Why it matters
Sora is more than a flashy experiment — it’s a glimpse at how AI video could become social infrastructure, not just a creative tool. By combining personal likeness, collaborative features, and real safeguards, OpenAI is positioning Sora as both playful and responsible.
At launch, the Sora app is only available by invite in the U.S. and Canada but the Sora 2 model is accessible worldwide for ChatGPT Pro users.
Stay curious,
Kinomoto.Mag
KINOMOTO.MAG 