OpenAI has announced the closure of the Sora app, which initially garnered significant audience attention. The official reasons for this decision have not been disclosed.
We’re saying goodbye to the Sora app. To everyone who created with Sora, shared it, and built community around it: thank you. What you made with Sora mattered, and we know this news is disappointing.
We’ll share more soon, including timelines for the app and API and details on…
— Sora (@soraofficialapp) March 24, 2026
Shortly after the launch of the updated video generator Sora 2 and its accompanying social network, the product caused a stir, despite being accessible only by invitation. The number of downloads reached 1 million in less than five days—faster than ChatGPT did in its time.
Despite a strong start, OpenAI failed to maintain user interest. A sustained interest in the AI feed never materialized.
The Sora social network was conceived as a TikTok analogue with a focus on artificial intelligence. Its key feature, cameos, allowed users to scan their faces, create realistic deepfakes of themselves, and share them with others for collaborative video creation.
In its early stages, Sora was marked by weak moderation. The feed was filled with controversial videos featuring OpenAI CEO Sam Altman. Soon, deepfakes of well-known figures like Martin Luther King Jr. and Robin Williams appeared.
Subsequently, users shifted to creating content with copyright-protected characters.
The initial hype quickly subsided. In November, the app’s metrics peaked at around 3.3 million downloads in the App Store and Google Play. However, by February, the number of downloads had dropped to 1.1 million.
Nonetheless, the closure of Sora does not mean the disappearance of the technology. The Sora 2 model remains available, albeit now for a fee, within ChatGPT.
In mid-March, media reported OpenAI’s intention to integrate Sora into ChatGPT.
Refocusing
Earlier, OpenAI discontinued the Instant Checkout feature, launched in September 2025. This tool allowed users to confirm orders, delivery, and payment without leaving the chatbot interface.
“OpenAI underestimated how complex implementing transactions would be. On one hand, it’s somewhat surprising, but on the other, it’s not easy for retailers either,” noted Bob Hetu, an analyst at the research and consulting firm Gartner.
Instead of direct sales, the startup announced new product search capabilities in ChatGPT, emphasizing the clarity of the process.
“You can browse products, compare options in a single interface, and get detailed, up-to-date information—all in one place,” the blog states.
The company expanded the Agentic Commerce Protocol (ACP), adding support for product searches. As a result, ChatGPT receives more comprehensive, relevant, and current information.
The new service allows users to browse and compare products, upload images to search for similar items, and refine results in dialogue mode. The system’s speed has been increased internally.
Updates will be available to all ChatGPT Free, Go, Plus, and Pro users by the end of the current week.
In March, OpenAI released two new AI models: GPT-5.4 mini and GPT-5.4 nano. They are specifically optimized for tasks requiring minimal latency in response generation.