Breaking news! OpenAI shuts down Sora, Disney's $1 billion deal is canceled

Author | Jiang Yu                   Editor | Bing Qian

Sora has shut down, and the stunning “Sora Moment” that captivated the world has come to an end.

According to Zhidx.com on March 25, early this morning, OpenAI’s video generation product Sora announced its closure via its official account, shutting down its video generation app and Sora 2 model API service. At the same time, the $1 billion Disney partnership also fell through.

This product, almost regarded as a synonym for AI video generation, has now hit pause.

Once the news broke, the community quickly bid farewell. Many netizens shared a “Sora memorial image,” showing a group gathered around a tombstone with “Sora” engraved on it, alongside the logo of a competitor, Seedance 2.0, with obvious teasing. Just yesterday, Seedance 2.0 officially launched worldwide.

Some users expressed regret, saying “Sora had great potential.”

Well-known AI blogger NIK simply stated “As predicted,” and reposted an earlier post predicting “Sora will disappear,” claiming its ending was long foreseen. In a previous graph showing download trends, Sora’s popularity steadily declined.

Official information indicates that the shutdown covers two core capabilities: one is the standalone Sora app for users, and the other is the Sora 2 model API for developers. Both ordinary creators and developers building video workflows based on Sora will lose direct access. The Wall Street Journal further reported that, besides the app and API, OpenAI will also cease supporting video-related features within ChatGPT.

The Sora team said they will announce the timeline for the app and API shutdowns and provide options for users to save and migrate their work.

Notably, this decision came quite suddenly. Just this week, OpenAI was still actively updating the Sora product.

Looking back, Sora first debuted in early 2024 and immediately caused a stir in the industry, regarded as a key milestone in AI video generation; later, in 2025, it launched Sora 2, with iOS and Android apps. The iOS version once topped the App Store download charts, and the product evolved from a tool into an “AI version of TikTok.”

However, the competitive window was short-lived. As models from Runway, Luma, Google Veo, and domestic players like Kuaishou KeLing, MiniMax Haijia AI, and ByteDance Seedance rapidly iterated, Sora’s leading advantage was quickly matched or even surpassed, and user interest declined.

On the business side, things also changed. In December 2025, Disney invested $1 billion to license its IP for Sora video generation and planned to distribute related content on Disney+. However, according to foreign media reports, that partnership was canceled along with Sora’s shutdown. Disney stated it would continue exploring AI collaborations, provided they respect copyright and creator rights.

Deeper changes involve OpenAI’s overall strategic shift. Several foreign outlets previously reported that OpenAI is focusing resources on ChatGPT, aiming to develop a “super app” integrating chat, coding, and creative capabilities, with video generation once considered for inclusion.

But as resources and priorities shifted, video generation gradually was pushed out of the core focus.

The Wall Street Journal also mentioned that OpenAI executive Fidji Simo referred to projects like Sora as “side quests” in internal meetings, emphasizing that resources should return to core products like the Agent system. This focus is also believed to be related to OpenAI’s preparations for a potential IPO. Previously, internal concerns had been raised about the massive computational resources consumed by Sora.

OpenAI co-founder and CEO Sam Altman also revealed that the Sora team will shift toward long-term directions, such as robotics.

In the longer term, OpenAI is also pushing organizational restructuring and has committed $1 billion through its foundation to areas like life sciences, employment, and AI safety. Sora’s exit appears to be a conscious choice in this strategic realignment.

Conclusion: From a “nuclear launch” to being sidelined

Sora once defined the upper limit of AI video imagination with a “nuclear-level release,” but what truly influenced its trajectory were factors like computing power, product pathways, and company priorities. As resources increasingly shifted toward Agent, coding, and enterprise scenarios, Sora’s position changed. Under this internal adjustment, the product was ultimately shut down.

That globally stunning “Sora Moment” has now come to an end.