Sora is OpenAI's AI video generation model that creates realistic and imaginative video content from text prompts and image references. Originally launched in December 2024 and significantly upgraded with Sora 2 in September 2025, it represents one of the most advanced text-to-video AI systems available. Sora 2 dramatically improved upon the original by extending video length from 6 seconds to up to 25 seconds, introducing synchronized audio generation including dialogue, sound effects, and music from a single prompt, and adding character cameo capabilities through OpenAI's partnership with Disney. The model demonstrates strong understanding of context, spatial relationships, and physics, producing coherent video sequences with natural motion and lighting. Key capabilities include text-to-video generation for creating scenes from detailed prompts, image-to-video conversion for animating still images, and storyboard mode for multi-scene video creation. Sora is integrated into ChatGPT subscriptions, with Plus subscribers getting unlimited 480p access and Pro subscribers receiving up to 1080p resolution with 10,000 monthly credits. An API is also available for developers at usage-based pricing. Since January 2026, Sora access requires a paid ChatGPT subscription, with the free tier no longer supporting video generation.