OpenAI Shifts Focus, Ends Sora Development

▼ Summary
– OpenAI is discontinuing its Sora video app and API roughly six months after launch to refocus its efforts.
– The company is shifting strategy to prepare for an IPO, moving away from a bottom-up culture that spread resources thin.
– A key new focus is developing a “superapp” that combines ChatGPT, Codex, and Atlas into a unified consumer assistant.
– OpenAI is bolstering its enterprise business, with Codex becoming a revenue leader surpassing $1 billion annually.
– The Sora shutdown has damaged a partnership with Disney, which has canceled a planned $1 billion investment.
In a significant strategic pivot, OpenAI has announced the discontinuation of its Sora video generation application and its corresponding API. This decision, made just over half a year after the product’s launch, signals a major corporate refocusing as the company prepares for an anticipated initial public offering. OpenAI’s CFO, Sarah Friar, emphasized the need for the organization to streamline its operations, stating the company must be prepared for the scrutiny and demands of the public market.
This shift marks a departure from the company’s recent operational style. Under CEO Sam Altman, OpenAI has functioned with a bottom-up culture, empowering teams to pursue a wide array of ambitious projects. This approach, reminiscent of the Y Combinator incubator model, led to ventures into video generation with Sora, hardware devices, robotics, and its Codex coding agent. While fostering innovation, this expansive strategy has also stretched the company’s computational resources and talent thin, with some initiatives failing to gain sustained traction.
The data underscores this challenge for Sora. After a strong launch that saw 3.3 million global downloads in November 2025, interest sharply declined, plummeting to just 1.1 million downloads by February 2026 according to analytics from Appfigures. This stalled growth contributed to the decision to reallocate Sora’s dedicated researchers and powerful GPUs to other priority areas.
Leadership has now issued a clear mandate to concentrate efforts on a narrower set of core objectives. A primary focus is the development of a unified AI superapp that integrates ChatGPT, Codex, and the Atlas project into a single consumer interface. This initiative revives the original vision of creating a true “super assistant” capable of handling complex digital tasks, a goal that has proven more difficult than initially anticipated. By consolidating these tools, OpenAI aims to create a more compelling and widely adopted agent for its user base.
Concurrently, the company is aggressively strengthening its enterprise business segment. The Codex team has become a particular success story, having caught up to early leaders in the AI coding space. Codex now represents a major revenue driver, surpassing $1 billion in annualized revenue earlier this year and continuing its expansion.
The winding down of Sora has not been without consequence. It appears to have abruptly ended a high-profile partnership with Disney, which had previously committed to a $1 billion investment in OpenAI. Reports indicate Disney was caught off guard by the decision and has subsequently withdrawn its investment plans.
Internally, this new era of focus presents both opportunity and risk. While many employees are reportedly energized by the clearer direction, the competition for top AI research talent remains fierce with rivals like Anthropic, Google DeepMind, and Meta. The recent departure of VP of Research Jerry Tworek, who struggled to secure resources for a new project, highlights the potential for talent attrition if other long-term research initiatives are deprioritized. As for the Sora team itself, OpenAI states its members will transition to world simulation research aimed at advancing robotics for solving physical tasks, aligning with the company’s refined strategic goals.
(Source: Wired)




