Pradeep Viswanathan Neowin @pradeepviswav ·
Microsoft's Azure AI Foundry provides developers access to more than 1,900 AI models suitable for a variety of use cases and industries. However, Azure AI Foundry currently lacks text-to-video generation models, which are already available on AWS via its Nova Reels API and on Google Cloud via its Veo API.
Microsoft relies on OpenAI for all its frontier AI needs. Although OpenAI has developed the Sora video generation model, it has not yet been made available as an API for developers. However, ChatGPT Premium plan customers can use Sora to create videos using text prompts. Due to the significant GPU resources required to make Sora available to all ChatGPT users and developers, OpenAI has not yet rolled it out widely, despite launching it back in December 2024.
At Build 2025, Microsoft announced that Sora will be available in Azure AI Foundry starting next week. The company also introduced a new 'Video Playground' section within Azure AI Foundry, where developers can experiment with video generation models like Sora. Developers will be able to customize the aspect ratio, resolution, and duration of the generated videos. Additionally, Sora will be accessible via the Azure OpenAI Service, allowing developers to integrate it into their applications.
— Microsoft Events (@events_msft) May 19, 2025Sora in Azure AI Foundry will supercharge creative workflows to enhance storytelling and deliver more personalized experiences. ?
Coming soon, as part of the trusted network of agents Microsoft is creating with partners like @OpenAI. pic.twitter.com/R5zC5t2WXk
Microsoft also highlighted that T&Pm, a WPP company, is already using Sora through the Azure OpenAI Service to enhance its workflow. The company is leveraging Sora to easily visualize early concepts and scale big ideas through to production. You can read more such use cases for the Sora API here.
