OpenAI Shuts Down Sora: $15M Daily Costs and Collapsed Disney Deal

The End of OpenAI's Video Ambition

On March 24, 2026, OpenAI announced that it is shutting down Sora, its AI-powered video generation application. The decision comes just six months after the app's launch in September 2025 and marks one of the most high-profile product discontinuations in AI industry history. OpenAI stated simply: "We're saying goodbye to Sora," promising to share timelines for preserving user content and sunsetting the API.

The shutdown is driven by a stark financial reality: Sora was burning through an estimated $15 million per day in inference costs while generating only $2.1 million in total lifetime revenue from in-app purchases. Bill Peebles, OpenAI's head of Sora, acknowledged that "the economics are completely unsustainable."

The Numbers Behind the Collapse

Sora's trajectory from launch to shutdown tells a clear story of declining interest and escalating costs:

The app initially generated significant excitement, becoming the most-downloaded app in the iOS Photo and Video category on its first day. But sustaining user engagement proved impossible, and the compute requirements for real-time video generation consumed resources that OpenAI could deploy more profitably elsewhere.

Disney Deal Collapse

The shutdown's most prominent casualty is a reported $1 billion investment and licensing deal with Disney, announced in December 2025. The partnership would have allowed Sora users to generate videos featuring more than 200 characters from Disney, Marvel, Pixar, and Star Wars properties.

With Sora's discontinuation, Disney confirmed that no money was exchanged before the deal was terminated. A Disney spokesperson stated that the company "respect[s] OpenAI's decision to exit the video generation business." The collapse represents a significant missed revenue opportunity and a public setback for OpenAI's content creation ambitions.

Strategic Reallocation

OpenAI framed the shutdown as a strategic decision rather than a failure. The company stated that the Sora research team "continues to focus on world simulation research to advance robotics," suggesting that the underlying video generation technology will be redirected toward training robotic systems rather than consumer-facing content creation.

The compute resources previously dedicated to Sora will be reallocated toward what OpenAI considers more profitable ventures: coding assistance, reasoning capabilities, and text-generation tasks. This aligns with a broader strategic refocusing led by Fidji Simo, OpenAI's CEO of Applications, who recently acknowledged that the company had "spread its efforts across too many apps" and needed to simplify.

Competitive Pressure

The decision also reflects competitive dynamics in the AI industry. While OpenAI invested heavily in video generation, Anthropic's Claude models gained substantial traction among businesses and developers by focusing on text and code generation. The contrast highlights a strategic trade-off: video generation requires enormous compute for relatively low monetization, while text and coding assistants command higher willingness to pay from enterprise customers.

Other video generation competitors, including Runway, Pika, and emerging Chinese platforms, continue to operate but face similar questions about long-term economic sustainability of AI video at scale.

Copyright and Safety Concerns

Sora's brief existence was also marked by controversy. Users generated videos featuring copyrighted characters like Lara Croft and Mario without authorization, raising intellectual property concerns. Deepfake experts warned about the potential for misuse, and the platform struggled to enforce content policies at scale.

These issues added regulatory and reputational risk to the already challenging economics, making Sora a liability on multiple fronts.

Impact on Existing Users

OpenAI has promised to provide details on preserving existing user content and work created with Sora. The company committed to sharing specific timelines for both the app and API shutdowns. For users who had integrated Sora into their workflows, the discontinuation requires finding alternative tools for AI video generation.

Outlook

Sora's shutdown is a cautionary tale about the economics of AI-powered media generation. While the technology itself remains impressive, the gap between what users are willing to pay and what it costs to run video generation at scale proved insurmountable. For OpenAI, the $110 billion in recent funding and approximately $730 billion valuation provide ample runway to absorb the loss and redirect toward higher-margin products.

The decision may also signal a broader industry correction: AI video generation as a standalone consumer product may not be viable at current compute costs. The technology's future likely lies in enterprise applications, visual effects workflows, and robotics training rather than direct-to-consumer social feeds.

Conclusion

OpenAI's shutdown of Sora after just six months represents one of the fastest and most expensive product failures in AI history. With $15 million in daily costs, declining user engagement, and a collapsed $1 billion Disney partnership, the decision was economically unavoidable. The underlying technology will live on in robotics research, but the consumer video generation dream is over, at least for now. Developers and creators who relied on Sora should begin migrating to alternatives like Runway or Pika while OpenAI finalizes its shutdown timeline.