OpenAI has announced the rollout of Sora Turbo, an advanced version of its text-to-video generation model.
The new model, which builds on February's preview, offers significant speed improvements and enhanced features, marking a step forward in AI's ability to simulate and interact with the physical world.
According to OpenAI, Sora Turbo enables users to generate realistic videos up to 1080p resolution and 20 seconds in length. Users can select from widescreen, vertical, or square aspect ratios and have the option to create videos entirely from text or incorporate their own assets for remixing and blending.
OpenAI said that uploads featuring human subjects will initially be limited to minimize risks, with plans to expand access as the company refines its deepfake mitigation systems. The firm has implemented monitoring systems, including internal tools and red-teaming efforts, to ensure responsible use.
Sora is included in ChatGPT Plus subscriptions at no extra cost, allowing users to generate up to 50 videos at 480p resolution per month or fewer videos at 720p.
OpenAI is also developing tailored pricing options, expected to launch early next year.