sora

OpenAI announced Tuesday it will shut down Sora, the video generation app that launched publicly in December 2024. The shutdown comes as OpenAI refocuses on business and productivity applications and faces intensifying competition from ByteDance's SeeDance 2.0 and Google's Veo.

OpenAI announced the discontinuation of Sora, its consumer video generation app and API. The shutdown follows declining user engagement and aligns with OpenAI's strategic pivot toward enterprise AI products and robotics research, with Disney exiting a planned $1 billion investment.

OpenAI announced Tuesday that it is discontinuing Sora, its video generation tool launched in late 2024, along with both the standalone app and developer API access. The shutdown also terminates Disney's $1 billion investment deal announced in December, which included licensing Disney characters for use within Sora and plans to distribute AI-generated videos on Disney+.

OpenAI plans to add its Sora video generation model directly to ChatGPT, according to The Information. The move comes as the standalone Sora app, launched in September 2025, has seen declining engagement despite strong initial adoption. Integration could help grow ChatGPT's 900 million weekly active users toward 1 billion.

OpenAI released version 2.27.0 of its Python SDK on March 13, 2026, adding significant improvements to the Sora video generation API. The update introduces character API support, video extension and editing capabilities, and higher resolution export options.

OpenAI plans to integrate its Sora video generator as a built-in feature within ChatGPT, according to The Information. Currently available only on a standalone website and app, the integration would let users generate videos directly in the chatbot, similar to how image generation was added last year.

An international research team released the largest video reasoning dataset to date—roughly 1,000 times larger than previous alternatives. Testing reveals that state-of-the-art models including Sora 2 and Veo 3.1 substantially underperform humans on reasoning tasks, suggesting the limitation isn't data scarcity but architectural constraints.