OpenAI has announced that it will shut down its Sora video generation platform. This includes both the Sora iOS app and the developer APIs. The company will share the exact timeline in the coming days.
Sora was first introduced in early 2024 as an AI model that could generate realistic videos from text prompts. After months of testing, it was rolled out to paid ChatGPT users later that year. In 2025, OpenAI expanded the platform with the Sora 2 model and a dedicated iOS app, which allowed users to create and remix videos.
The app saw strong early interest. It crossed 100,000 downloads on day one and later reached over one million downloads. However, its popularity dropped over time. By early 2026, the app was no longer among the top apps on the iOS App Store.
Now, OpenAI has decided to shut it down completely. This also means developers will lose access to Sora APIs that allowed them to build apps using AI video generation.
The main reason behind this move appears to be cost. Running AI video models requires huge computing power. Reports suggest that the cost of serving users and developers was too high compared to the revenue it generated.
At the same time, OpenAI is shifting its focus towards coding-related AI tools. This space is seeing strong demand as developers are willing to pay for tools that improve productivity. Competitors like Anthropic are already seeing major growth by focusing on coding AI.
This decision also impacts OpenAI’s partnership with The Walt Disney Company, which was exploring AI video generation use cases.
This move again validates the same point I raised many times. Not every AI idea turns into a sustainable business. Companies have been trying to push AI into every possible product, but real-world usage does not always match expectations.
Walmart recently scaled back its AI-driven commerce efforts with OpenAI due to low conversion rates. This shows that adding AI does not automatically improve user experience or business outcomes.
The same applies to customer support. Many companies are replacing human support with AI chat systems, but the experience is often frustrating. In many cases, users struggle to get real help, which makes the overall service worse.
AI is powerful, but it has limits. Not every problem needs an AI solution. Companies need to focus on where AI actually adds value instead of forcing it into every product.
