In February 2024, OpenAI introduced Sora on a page titled “Sora: Creating video from text.” Sora is a model that generates video clips directly from natural-language descriptions, extending the text-to-image capabilities of earlier systems into moving images.
Sora represented a significant step in generative media because video is far harder to produce than still images. It requires the system to keep objects, characters, and scenes consistent across many frames and to render plausible motion over time. OpenAI presented Sora as a demonstration that AI could synthesize coherent video from a prompt, opening a new modality for generative AI.
For business readers, Sora signaled that video, one of the most expensive and labor-intensive forms of media to produce, was entering the generative-AI era. The announcement set expectations for how AI might reshape marketing, entertainment, and communication, even as the technology continued to mature toward production use.
The February reveal was a preview, not a product. On December 9, 2024, OpenAI moved Sora out of research preview and made it generally available, releasing a faster version called Sora Turbo as a standalone product at sora.com for ChatGPT Plus and Pro subscribers in the United States and several other countries. Plus and Pro tiers differed in resolution, clip length, and the number of generations allowed, the usual shape of a capability moving from demo to paid offering. The roughly ten-month gap between the first jaw-dropping clips and a usable product underscored how much harder it is to ship reliable generative video than to demonstrate it once.
Note on sourcing: OpenAI’s pages return an HTTP 403 error to automated fetchers, so the December general-availability details were corroborated through search against the canonical openai.com/index/sora-is-here/ post.