Sora had a big launch. Now OpenAI wants to give it a much bigger audience.

According to a report from The Information, OpenAI plans to bring its Sora video generation model directly into ChatGPT. It’s a move that makes a lot of sense when you think about it. Sora launched with serious buzz, but that excitement faded fast. Now OpenAI wants to plug it into the app that already has nearly a billion weekly users.

Sora’s Popularity Dropped Fast

When Sora launched alongside Sora 2 back in September 2025, it felt like a genuine moment. The standalone app was a hit right out of the gate.

But the momentum didn’t last. Users kept bumping into limits on how many videos they could create and what kinds of content they could generate. That friction took a real toll. The Sora app has since dropped out of the App Store’s top 100 free apps, and only a small fraction of users actually share videos publicly in the app.

Sora video generation integrated directly into ChatGPT interface

So the standalone experience, while technically impressive, never quite found its groove with everyday users.

ChatGPT Could Give Sora a Second Life

Here’s where it gets interesting. ChatGPT reported 900 million weekly active users as of February 2025. OpenAI wants to push that number past one billion, and adding video generation directly into ChatGPT could help get there.

Bringing Sora into ChatGPT puts the video tool in front of an audience that’s already engaged, already comfortable with the interface, and already paying for Plus or Pro subscriptions. That’s a very different situation than asking someone to download a separate app just to try video generation.

For context, the standalone Sora app isn’t going away. According to The Information, it sticks around even after the integration happens. But the real action will likely shift to wherever the bigger audience already lives.

Sora video generation model integrated directly into ChatGPT interface

The Cost Problem Is Very Real

This is where things get complicated. Video generation is expensive, and scaling it to ChatGPT’s audience is a genuinely big financial challenge.

OpenAI currently charges API customers $0.10 per second for a 720p video. During Sora’s launch period, the company offered 30 free video generations per account per day. That was already generous. Now multiply that generosity across hundreds of millions of ChatGPT users, and the numbers get uncomfortable quickly.

The Information reports that OpenAI has projected it could spend over $225 billion on inference — the cost of actually running its models — between 2026 and 2030. Inference costs cover every query, every image, every video. Adding Sora to ChatGPT will only push that number higher.

Credits and Disney Characters Could Be the Answer

So how does OpenAI make the math work? The credit model is the most obvious path forward.

OpenAI inference costs and credit model for scaling Sora video generation

OpenAI already tested a credits system inside the Sora app, requiring users to pay for additional generations once free allowances run out. Something similar will likely show up inside ChatGPT. You might get a handful of free video generations per month, then pay credits or upgrade your subscription to unlock more.

And there’s one detail from The Information worth noting. OpenAI’s partnership with Disney could play a role here. The ability to generate videos featuring Disney characters might become a premium perk that tempts users into buying more credits once their free generations run dry. It’s a clever way to make the credit wall feel like an opportunity rather than an obstacle.

A Bigger App, a Bigger Bill

Whether Sora inside ChatGPT becomes a must-have feature or just a novelty people try once, OpenAI’s costs are going up either way. The company is betting that growing ChatGPT’s active user base past a billion is worth the spend. And with competitors adding video generation tools of their own, standing still isn’t really an option.

The real question isn’t whether Sora deserves a second chance. It probably does. The question is whether users will embrace video generation as a regular part of their ChatGPT workflow, or treat it as a fun experiment they forget about after a week. OpenAI is spending a lot of money on the hope that it’s the former.