From Center of Attention to a Hasty End: Why Sora Became OpenAI’s "Discarded Piece"

From Center of Attention to a Hasty End: Why Sora Became OpenAI’s "Discarded Piece"

```

OpenAI once positioned the video generation tool Sora as the most important consumer-level product following ChatGPT, but now it has abruptly ceased operations less than six months after being opened to the public. This outcome perhaps indicates that, in the AI race where computing power is scarce and competition is intensifying, products that keep burning money without monetizing have no place.

The direct trigger for Sora's shutdown was the harsh business ledger—global active users plummeted from a peak of about one million to less than 500,000, with daily operational losses around one million dollars. Meanwhile, OpenAI is racing to free up computing power for its new model codenamed "Spud," to support programming and enterprise service product development, while Sora is consuming too much of this most precious resource.

This decision directly impacted OpenAI’s core partner. Disney had previously signed a multi-year licensing agreement with OpenAI, pledging $1 billion in investment and granting access to over 200 of its IP characters for Sora. Many Disney executives learned of the shutdown decision less than an hour before it was announced; the $1 billion investment has yet to be realized, and the collaboration has effectively stalled.

OpenAI stated it will concentrate the freed computing power into productivity tools to catch up with its competitor Anthropic, which has taken the lead in the enterprise market. This shift shows that, at a critical IPO juncture, the ability to commercialize has become the primary criterion for resource allocation.

Ephemeral Success: The Growth Dilemma Behind Surging Popularity

Sora was born from the academic ideals of two UC Berkeley PhDs. Tim Brooks and Bill Peebles joined OpenAI in early 2023, aiming to build an AI model capable of generating high-quality videos from text and simulating the physical world. In February 2024, they named the system Sora with the Japanese word for “sky” and debuted it to the public. Sora shocked the industry with realistic videos like mammoths traversing snowy plains and stylish women strolling Tokyo's neon-lit streets; Sam Altman promptly invited users on the X platform to submit text prompts and showcase its generation capabilities.

In December of the same year, OpenAI officially launched the Sora consumer app to the public. Within a week, the app shot to the top of the App Store rankings. Users simply input a prompt to receive a ten-second short video in minutes; it also supported uploading one's own face, allowing users to appear in various surreal scenes. Altman himself contributed his own likeness, sparking a wave of curiosity among users.

However, the hype faded just as quickly. After reaching about one million global users at its peak, the user base continuously declined, shrinking to less than 500,000 in the following months. According to data analytics firm Similarweb, usage had stagnated by year-end. Altman once compared Sora's launch to ChatGPT’s historic debut, but the app ultimately failed to realize creators’ ambitions—The Wall Street Journal described its performance as "more like AI junk than AI magic."

Computing Power Black Hole: The Cost of a Million Dollars Per Day

The high operating costs of Sora stem from the technical nature of video generation models. Unlike language models that learn from text, video models must understand and reconstruct entire dynamic scenes, making their training and inference costs much higher than text-based products. Every user who embeds their face into a WWII newsreel or Hollywood chase scene is consuming a share of OpenAI's finite computing power. According to media reports citing insiders, Sora was losing around one million dollars every day.

An internal dashboard tracked computing power allocations among OpenAI teams. Some employees were surprised at the allocation Sora received—since the video generation tool could not bring considerable income or boost the language model’s capacity, this allocation sparked internal doubts.

External competition made the problem even more urgent. Google's Gemini earned widespread consumer adoption; Anthropic’s Claude Code quickly captured the software engineering community in Silicon Valley thanks to its autonomous programming abilities, catching OpenAI off guard. OpenAI then hastily released a new version of its own programming product Codex, but still struggled to close the gap.

The new model codenamed "Spud" urgently needed more computing power. The company also planned to train a new model specifically for video generation features in ChatGPT. After accounting costs, OpenAI ultimately decided to cancel this training plan and completely shut down Sora.

Meta Poaching and Internal Isolation

Sora once nearly ended early due to a talent war. In spring 2025, Meta CEO Mark Zuckerberg launched a massive poaching campaign targeting OpenAI, personally reaching out to dozens of top researchers with lucrative offers to join Meta's AI lab. Sora co-founder Bill Peebles was among those targeted and seriously considered the offer. According to media reports citing insiders, OpenAI retained Peebles by raising his salary and expanded his responsibilities on the Sora project, allowing him to lead next-generation video model training and consumer app development.

However, Sora always operated in a highly isolated state within the company. The project belonged to the world simulation team led by Aditya Ramesh, operating independently from the core research team responsible for ChatGPT’s language models, and its progress was kept highly confidential from other departments. Some departing employees described Sora as "a company within the company." This isolated mode to some extent made it difficult for Sora’s strategic value to gain wider internal consensus, leaving it relatively vulnerable in resource competition.

Disney's Broken Dream: $1 Billion Investment Turns to Nothing

Sora’s most prominent partnership ended just as abruptly.

In December 2024, OpenAI and Disney announced a multi-year licensing agreement covering more than 200 famous IP characters, including those from Marvel and Pixar. Disney pledged to invest $1 billion and become OpenAI’s major enterprise client. Disney CEO Bob Iger told CNBC the cooperation let Disney participate in AI’s rapid growth and shape new forms of media entertainment. Notably, the day before announcing the deal with OpenAI, Disney had sent Google a cease-and-desist letter for copyright infringement; this was seen as an endorsement of AI licensing business models.

In February this year, Iger revealed on an earnings call that Sora-generated short videos would soon appear in Disney+’s vertical video feed, and that negotiations were underway to introduce ChatGPT across Disney company-wide.

However, many Disney executives learned of Sora’s closure less than an hour before OpenAI made the announcement. The $1 billion investment has not been realized, and the partnership is effectively stalled. Disney then issued a statement: "With the rapid evolution in AI, we respect OpenAI’s decision to exit video generation and focus on other directions, and appreciate productive collaboration between the teams." Under new CEO Josh D'Amaro, Disney is currently actively negotiating with more than ten partners about plans to introduce other AI tools.

Betting on the "Super App": Pragmatic Logic Takes Over

Sora’s shutdown is not just a product failure, but a microcosm of OpenAI’s rapidly narrowing strategic focus on the eve of IPO.

Altman described the decision in an internal staff memo as "a difficult but necessary sacrifice for the company’s overall objectives," expressing appreciation for employees’ willingness to make "tough trade-offs." A female OpenAI spokesperson said the company was "ruthlessly prioritizing" computing resources based on maximizing long-term economic value, saying "this focus allows us to grow and innovate more quickly and to serve enterprises and developers more efficiently."

The company is currently shifting focus to a planned "super app" that will integrate so-called "agents" AI tools, enabling them to autonomously perform tasks like coding, data analysis, and itinerary booking for users. Productivity-oriented products like this are rapidly spreading among enterprise and developer groups, but OpenAI is currently lagging behind Anthropic in this market. Altman said the former Sora team would now redirect its efforts toward robotics and other areas with greater long-term potential.

For OpenAI, Sora's journey was a costly strategic misjudgment. Altman once hoped the product would reshape pop culture and open up new sources of revenue, but ultimately it failed due to lack of a viable business model and improper resource allocation. At the key moment when the company is focusing on profitability, Sora became a burden that had to be removed.

Risk Warning and DisclaimerMarkets are risky. Investments must be cautious. This article does not constitute personal investment advice, nor does it take into account an individual user's particular investment goals, financial situation, or needs. Users should consider whether the opinions, views, or conclusions in this article are suitable for their specific circumstances. Investing based on this information is at your own risk. ```