OpenAI Shuts Down Sora and Winds Down Separate Video App After Viral Launch

OpenAI is closing Sora — an app for video generation on iPhone that seemed like the company's next viral hit back in fall 2025. The experiment with a separate video service lived briefly: the hype turned out to be stronger than actual demand, and subsequent restrictions quickly narrowed down usage scenarios.

How Sora took off

Sora launched in September 2025 and became OpenAI's second iPhone app after ChatGPT. The idea was simple: give users the easiest possible path from a text prompt to a short video. The service initially worked almost like a viral content constructor — it generated videos of different genres, required no complex training, and came with a built-in feed where you could watch and share others' work.

For a mass audience, this made Sora less of a video editing tool and more of a blend between video editor and social platform. It was precisely this simplicity that quickly brought the service attention. Sora was discussed as an app capable of turning video generation into a familiar mobile scenario: open, type a couple of lines, get a video, post it to the feed.

But such growth had a weak side. Virality doesn't always mean retention, and interest in "look what the neural network can do" rarely lasts long if real daily value doesn't emerge behind the effect. Without it, Sora turned from habit into an attraction.

Why everything was shut down

Problems began when OpenAI started stricter restrictions on using third-party intellectual property. The initial freedom of generation was one of the reasons for Sora's popularity: users came for the ability to quickly assemble recognizable scenes, styles, and meme videos. When these boundaries narrowed, the magic of "do whatever you want" disappeared. As a result, the service ended up between two audiences: for professionals it didn't have time to become an essential tool, and for the general public it ceased to be daring and unpredictable enough.

"We're saying goodbye to Sora.

What you did with it mattered," — the Sora team.

• Viral launch after release
• Simple video generation from text
• Restrictions on content using third-party intellectual property
• Weak conversion of hype into sustained use

OpenAI has already confirmed the shutdown in a message on X and promised to later explain what will happen with the app's shutdown timeline and API, as well as how users will be able to save their work. This is an important detail: the company isn't just removing a mobile shell, but essentially closing an entire direction that was presented as the next notable step after text and voice products. For authors testing Sora as a separate platform, this means the need to quickly find an alternative route.

Where focus is shifting

The story may not end with just closing one app. According to media reports, OpenAI is winding down broader efforts around generative video and redistributing resources in favor of a "super app." The logic is clear: instead of a set of separate services, the company wants to build a single product where ChatGPT, development tools, and browser functions work as one system.

With such a strategy, a separate video platform no longer looks like a priority but rather an experiment that didn't achieve the needed scale. This also changes the assessment of Sora retroactively. The project was often perceived as a vivid demonstration of the model's capabilities, but not as a mature professional tool on which you could build a sustainable workflow.

Interest in it was shown by major players: the article mentions potential collaboration with Disney, which ultimately didn't result in a real stress test. If the company truly is moving away from an independent AI video product, it means even a notable brand and strong initial attention proved insufficient without a clear business model and long-term retention.

What this means

Sora's closure shows that on the AI market there's little value in simply surprising users with "wow" content generation. The winners are not the most viral demos, but products that retain an audience, fit into daily tasks, and withstand copyright restrictions. For OpenAI, this is a pivot toward an ecosystem around ChatGPT; for the market, it's a reminder that the era of separate AI toys is ending faster than it seems. This applies to both mobile apps and standalone content generators.