OpenAI Closes Sora: A Major Strategic Shift

OpenAI Ends Sora, Its Video Revolution

In February 2024, OpenAI shook the tech world with the launch of Sora, a video generation model that allowed users to create nearly cinematic-quality videos from simple text instructions. This model was one of the first of its kind to be accessible to the general public, forcing competitors to accelerate their efforts to keep up. In September 2025, Sora 2 was introduced as a standalone application, offering even more realistic videos accompanied by compelling soundtracks. For many, OpenAI seemed to have achieved perfection with this product, especially after signing a partnership with Disney in the fall of 2025, allowing the use of licensed characters in the generated videos.

However, yesterday, the company announced the shutdown of Sora without providing any real public explanations, merely posting a message on X. “We bid farewell to the Sora app. To everyone who created with Sora, who shared it, and who built a community around it: thank you… We will share more soon, including the timeline for shutting down the app and API, as well as how to preserve your work.” This sudden announcement caught everyone off guard.

Exorbitant Costs for Meager Revenues

According to Forbes, the operating costs of Sora reached staggering heights, with expenses potentially soaring to $15 million per day, amounting to over $5 billion annually. In contrast, the revenue generated by the app only reached $1.4 million since its launch, with a monthly peak of $540,000. Bill Peebles, head of Sora at OpenAI, had already admitted in October 2025 that the business model was “totally unsustainable.”

The computing power required for each generated video strained the resources available for other teams at OpenAI. The produced clips, while technically impressive, did not provide any real added value to the company, failing to justify the astronomical sums spent. Moreover, the initial excitement for the app had waned: since early 2026, downloads had been declining month after month, with a 32% drop in December 2025, a period typically favorable for apps. OpenAI realized that offering a free movie studio to the entire world was akin to financial suicide and decided to close what might have been one of the most irresponsible chapters in its history.

Strategic Refocusing Before Going Public

The shutdown of Sora is part of a broader strategy to reposition OpenAI. Some had speculated about integrating the video model into ChatGPT, a way to salvage the technological foundation of Sora by merging it into the company's flagship product. However, Sam Altman made a decisive call: no. The application is dead, access for developers is gone, and no resurrection is expected.

The teams that worked on Sora have been redirected to OpenAI's robotics division. The goal is to concentrate resources on other priority research areas, particularly the development of artificial general intelligence (AGI), Sam Altman's ultimate dream.

OpenAI seems to want to follow the example of Anthropic, which has built an empire with its chatbot Claude, generating $19 billion annualized by early 2026, without ever offering its clients a tool to create images or videos. Anthropic has focused from the beginning on one thing: being irreplaceable for developers and businesses. Claude Code, its programming assistance tool, has become one of the market's benchmarks in this segment. This is precisely where OpenAI now wants to concentrate its efforts.

This will lead to a thorough restructuring of all its products: ChatGPT, Codex (its programming assistance tool), and ChatGPT Atlas, its AI-powered browser, may eventually merge into a single application.

Considering that a potential IPO is on the horizon by the end of the year, OpenAI knows exactly what this entails: opening its books to investors. In this regard, presenting a plethora of loss-making products is not exactly a good strategy, and Sora was the burdensome branch that needed to be pruned before showing a clean slate.