Burn Rate Apocalypse: OpenAI Pulled the Plug on Sora After It Was Losing Up to $1 Million a Day

OpenAI killed one of the most hyped AI products in recent memory on March 24, 2026, and the math behind the decision is staggering. Sora — the generative video app that topped the App Store, rattled Hollywood, and secured a billion-dollar Disney deal — was shut down after a Wall Street Journal investigation revealed it was hemorrhaging roughly $1 million every day in compute costs, against total lifetime consumer revenue that barely cleared $2 million. That is not a business model problem. That is a physics problem.

The numbers tell the whole story. When Sora's standalone app launched in September 2025, it became the most-downloaded Photo and Video app on the iOS App Store within 24 hours. Monthly active users peaked at around one million before collapsing to fewer than 500,000 — and the app was burning through roughly $1 million every day, not because people loved it, but because video generation is so costly to run. Every 10-second clip cost OpenAI an estimated $1.30 in compute, according to analysis by Cantor Fitzgerald. Scale that across millions of daily free users, and the arithmetic becomes irreversible.

The human cost of that math hit Disney the hardest. The entertainment giant had committed $1 billion to the partnership, yet found out Sora was being shut down less than an hour before the public announcement. The deal died with it. Disney — which had planned to bring over 200 characters from Marvel, Pixar, and Star Wars into the Sora ecosystem, with curated content eventually appearing on Disney+ — issued a statement confirming the exit.

Beyond the Disney fallout, the shutdown signals something bigger about where the AI industry is headed. The closure of the resource-intensive app comes ahead of an expected initial public stock offering from OpenAI in the coming months, with the company now diverting efforts from disparate consumer products toward business clients. OpenAI, valued at $730 billion following a $110 billion funding round in early 2026, cannot afford for its IPO prospectus to include a line item burning nine figures annually on a product with minimal revenue.

OpenAI's official statement cited a strategic focus on other priorities: "As we focus and compute demand grows, the Sora research team continues to focus on world simulation research to advance robotics that will help people solve real-world, physical tasks." Translated from corporate, that means: the chips that powered Sora videos will now power more profitable enterprise AI products, coding tools, and robot navigation systems.

The competitive pressure was equally decisive. While a whole team inside OpenAI was focused on making Sora work, rivals — including Anthropic with its Claude models — were quietly winning over the software engineers and enterprises that drive real revenue. By Q1 2026, Google Veo and Runway had also matched Sora's output quality while delivering results in a fraction of the time, eliminating what little competitive advantage remained.

Sora's collapse is, ultimately, a case study in the gap between technological capability and commercial durability. The app was genuinely impressive. It was also structurally unprofitable from day one, relying on consumer novelty cycles that wore off in roughly twelve weeks. UC Berkeley researcher Hany Farid, who specializes in digital imagery, noted that Sora's greatest legacy may be the spread of AI-generated video content that the public struggles to distinguish from real footage — a phenomenon that will outlive the app itself.