On a Tuesday morning in March, Disney executives arrived at OpenAI’s offices expecting another day of collaboration. They were working with Sora engineers on integrating Mickey Mouse, Marvel superheroes, and Star Wars characters into the AI video platform. Hours later, they learned the truth: Sora was dead.
The $10 billion partnership evaporated. The three-year licensing deal covering over 200 iconic characters vanished. And OpenAI’s once-celebrated “child” - the AI video model that stunned the world in 2024 - was being put down.
What happened to Sora wasn’t a slow decline. It was a sudden, brutal termination. And the reasons reveal something uncomfortable about the AI industry’s future.
The Rise and Sudden Fall of a Star
When OpenAI first showed Sora in early 2024, it felt like magic. Type a sentence, and a Hollywood-grade video appeared - whales flying through clouds, samurai wandering neon cities, scenes that looked ripped from a director’s imagination.
By September 2025, Sora had its own app. Downloads exploded: 4.8 million in October, 6.1 million in November. OpenAI CEO Sam Altman called it the “ChatGPT moment” for video.
Then came the crash. December downloads dropped to 3.2 million. January: 2.1 million. February: 1.4 million. By March, just 1.1 million.
The numbers told a brutal story. Sensor Tower data showed Sora’s 60-day user retention rate hit zero percent. People tried it once, said “wow,” and never came back.
The Numbers That Killed Sora
Here’s what OpenAI’s accountants saw when they looked under the hood.
Daily operating cost: Approximately $15 million. Some estimates put Sora’s annual burn rate at $5.4 billion.
Total revenue since launch: Around $1.4 million. Monthly peak revenue: $540,000 .
For every dollar Sora brought in, it burned thousands. A single 10-second video cost $1.30 to $3 in compute - before counting overhead.
Meanwhile, ChatGPT was generating $19 billion annually. The contrast was impossible to ignore.
Read More: Apple Will Hide Your Email Address From Apps and Websites, But Not Cops: What You Need to Know?
The Disney Deal That Never Was
In December 2025, OpenAI announced a blockbuster partnership with Disney. The entertainment giant would invest $10 billion. In exchange, Sora users could generate videos with Mickey Mouse, Darth Vader, Marvel heroes - over 200 iconic characters .
It seemed like a turning point. Hollywood’s biggest studio was legitimizing AI video.
But the deal never closed. Sources say the investment was structured as warrants, not cash. When OpenAI decided to kill Sora, Disney learned about it hours before the public announcement .
Disney’s official response was diplomatic: “We respect OpenAI’s decision to exit the video generation business” . Behind closed doors, executives were blindsided.
The Deeper Problem: A Toy, Not a Tool
Why did users abandon Sora so quickly? Because it couldn’t do real work.
Professional video editors need precision. They need character consistency across scenes. They need layered outputs, alpha channels, camera control. Sora gave them beautiful randomness .
You could prompt “a cat drinking coffee in a steampunk cafe” and get something amazing. But try to make a 30-second ad with a consistent brand look, and Sora fell apart.
It was a demo machine, not a production tool. People used it once, shared the result on social media, and never opened it again.
The IPO Pressure Cooker
OpenAI is preparing to go public. The company’s valuation has soared past $700 billion. But the losses are staggering: $80 billion in 2025, projected to hit $250 billion in 2026 .
Investors want to see profit, not experiments. And Sora was an expensive experiment that showed no path to profitability.
At an internal all-hands meeting, OpenAI’s apps chief Fidji Simo told employees: “We cannot miss this moment because we are distracted by side quests” .
Sora was a side quest. Codex - the AI coding tool competing directly with Anthropic’s Claude - was the main event. OpenAI redirected Sora’s computing resources to Codex, where paying customers actually existed.
Read More: Claude vs. ChatGPT vs. Gemini: The Winner Isn't Who You Think
What Happens Now
Sora’s team isn’t fired. They’re being reassigned to robotics and “world simulation” - technology that can help AI understand physics, not just generate pretty pixels .
OpenAI has also killed its instant shopping feature and is consolidating its browser, ChatGPT, and Codex into a single “super app” for productivity .
The message is clear: OpenAI is done chasing consumer hype. It’s going after enterprise dollars.
For the AI video industry, Sora’s death is a wake-up call. Chinese competitors like ByteDance’s Seedance 2.0 are thriving by focusing on short dramas - a market with clear monetization through subscriptions and ads . They proved that AI video can make money. Sora proved that cool tech alone isn’t enough.
Conclusion
Sora was OpenAI’s most beautiful failure. It wowed the world, landed a $10 billion partnership, and then collapsed under the weight of its own costs.
The lesson isn’t that AI video is dead. It’s that the era of “build it and they will come” is over. In 2026, investors want profits. Enterprises want tools. And consumers? They tried Sora, said “cool,” and moved on.
OpenAI chose to live another day. Sora did not.
FAQ
Q: Did OpenAI lose the $10 billion Disney deal because of Sora’s shutdown?
A: Yes. The deal was contingent on Sora’s continued operation. When OpenAI shut down Sora, the Disney partnership ended. The investment was structured as warrants, so no cash changed hands .
Q: How much money did Sora lose?
A: Estimates vary, but daily operating costs reached $15 million, with an annual burn rate around $5.4 billion. Total revenue from the app was about $1.4 million .
Q: Why did users abandon Sora so quickly? A: The 60-day retention rate was 0%. People tried it once, found it couldn’t produce consistent, usable results for professional work, and never returned.
Q: Is OpenAI completely abandoning video generation?
A: Yes for consumer products. The company is redirecting Sora’s research team to robotics and world simulation - technology focused on physical understanding, not entertainment.
Q: What happens to Sora’s competitors?
A: Companies like ByteDance (Seedance 2.0) are thriving by targeting specific markets like short dramas, where monetization is clearer. Google’s Veo also remains in the race.
i Personally felt that Sora didn't produce quality videos
ReplyDelete