OpenAI announced the closure of Sora, its AI video creation app and social platform. The decision also includes the end of access for developers to the Sora 2 series of video models, via application programming interface (API), which prevents incorporation into products or content generation processes.
Official communication occurred unexpectedly this week, through a post on OpenAI’s X (formerly Twitter). The company did not detail a precise date for the end of services, but promised to release “timelines for the application and API, as well as details on how to preserve its work.”
When opening the Sora app today, users are confronted with a farewell message in video format, created by AI. The presentation is made by exclusive Sora characters, developed by OpenAI, such as “littlecrabman”, and is accompanied by the following caption:
“An important update on Sora: Today, after careful internal discussion about our broader research priorities, we made the difficult decision to discontinue the Sora project. We know that many of you have invested significant time and energy in Sora – building not just generations, but also real audiences and communities. The creativity that has emerged has exceeded our expectations,” the message reads.
The statement ends by stating that the team responsible for Sora is “exploring ways to support the export and preservation of its work” for users.
More on this story: Microsoft confirms launch of Copilot super app to simplify digital experience in 2024
Sora had a meteoric rise and short life in the AI market
Sora made a huge global impact when it was unveiled by OpenAI in February 2024, more than two years ago, due to its ability to generate extremely realistic scenes. However, its updated version, Sora Turbo, released ten months later, was met with mixed reception. At that time, several other competitors in the field of video AI, such as Runway, Luma and Chinese companies Kling and Minimax, had already introduced notable rival models.
OpenAI, until then, demonstrated a strong commitment to improving the video model and offering resources to creators, launching Sora 2 via API and iOS applications, which quickly reached the top of downloads in the Apple App Store. A Sora app for Android was also planned for the second half of 2025. On both platforms, the goal of the Sora app was to function as a social network, allowing users to insert realistic versions of themselves and friends into videos created by artificial intelligence.
Interestingly, the company had released constant updates for Sora throughout this same week, which makes the announcement of its closure even more unexpected.
Disney cancels $1 billion deal with OpenAI after Sora decision
Sora had gained such market appeal that Disney, an entertainment giant, had pledged a $1 billion investment in OpenAI shares. This agreement was announced in December 2025, just four months ago.
The investment would allow the integration of popular Disney characters into Sora, enabling users to produce new videos with these figures and even insert themselves alongside them. The content generated would be distributed by Disney through its streaming service, Disney+, starting this year. OpenAI and Disney had jointly stated in 2025:
On the same topic: Google DeepMind innovation: new AI model controls entire robot body
“Under the license, fans will be able to watch selections of Sora-generated videos on Disney+, and OpenAI and Disney will collaborate to use OpenAI models to create new experiences for Disney+ subscribers, promoting innovative and creative ways to connect with Disney stories and characters. Sora and ChatGPT Images are expected to begin generating fan-inspired videos featuring licensed characters from various Disney brands in early 2026,” the statement said.
The cancellation of the contract with Disney was first reported by specialized vehicles, due to OpenAI’s decision to discontinue the Sora project. This information was later confirmed by other sources familiar with the matter. There are reports that the Sora and Disney teams collaborated intensely until just a few days before the announcement.
OpenAI’s focus shifts to robotics and real-world simulation
In an official statement, an OpenAI spokesperson detailed the reasons behind the company’s abrupt exit from the artificial intelligence video segment.
“We have decided to discontinue Sora in the consumer app and API. As our team focuses and demand for computing increases, the Sora research team continues to focus on real-world simulation research to advance robotics and help people solve real-world physical tasks,” the spokesperson explained.
Additionally, internal sources have indicated that OpenAI is redirecting its vast computing capacity toward the goal of achieving artificial general intelligence (AGI), a concept the company describes as AI systems capable of surpassing human performance in most activities of economic value.
Sora’s core technology, designed to understand physics, the physical environment and movement to create realistic videos, is now seen by the company as more valuable for applications in robotics and solving real-world challenges, rather than just a tool for generating media.
Learn more: Google DeepMind AI now with full control of humanoid robots in complex tasks
With the discontinuation of Sora, OpenAI signals a move away from the consumer entertainment and technology sector, moving towards areas with greater profit potential, such as manufacturing, logistics and automation in general. This change may be driven by a revenue and cost-benefit analysis, considering that the Sora app was free to use initially, while ChatGPT subscribers paid for additional features.
OpenAI advances in creating an integrated “super application”
The announcement about the end of Sora coincides with another revelation: a reorganization in the leadership of OpenAI and its philanthropic arm, the Foundation. The latter has pledged a $1 billion investment this year alone in areas such as life sciences, fighting disease, job creation, economic impact, AI resilience and community programs, indicating a clear shift in priorities regarding AI-generated content and media.
In parallel, OpenAI had already publicly expressed its objective of developing a “super application”, which would unify the functionalities of several of its products. These include the ChatGPT chatbot, the Codex AI coding model, the Atlas browser, and other tools, all in a singular interface.
Several market analyzes indicate that this strategic shift towards a Super App represents a move by the company to realign its focus and confront the growth of rivals such as Anthropic and its Claude line. Competition is notable in the enterprise and software developer segment, as Claude has seen rapid acceptance and enterprise usage in recent months, driven by its coding capabilities and automated execution of digital tasks.
Additionally, amidst the scenario of rising energy prices, exacerbated by the ongoing conflict between the USA, Israel and Iran, there is speculation that OpenAI seeks to adapt to control expenses. The discontinuation of Sora, a high energy consumption operation, would be part of this strategy, as video generation is intensely demanding on computational resources and energy, especially compared to the creation of text and code.
From this perspective, the end of Sora appears to be the result of a convergence of multiple distinct causes, many of them interconnected. They all point to the conclusion that producing videos for entertainment purposes no longer represents a priority investment of time and resources for OpenAI at this time.
However, this decision also raises questions about whether the company’s comprehensive and multifaceted approach, which aimed to master several fronts of AI simultaneously, had in any way harmed or limited its competitiveness against agile rivals. It remains to be seen whether OpenAI is now looking to correct its trajectory, or whether drastic changes in the artificial intelligence landscape, which favor competitors or new business opportunities, are forcing the company to completely recalibrate its strategy.