OpenAI introduced Sora generative AI to create photorealistic videos

By: Anton Kratiuk | 10.12.2024, 23:15

OpenAI has released a new Sora generative AI model for creating photorealistic videos.

Here's What We Know

Subscribers of paid ChatGPT Plus and ChatGPT Pro can already generate videos by text request.

For ChatGPT Plus users, this feature is limited to 50 pieces, 720p video quality, and a timeline of 5s.

Those who can spare $200 per month for ChatGPT Pro will get up to 500 videos with up to 1080p resolution and a maximum duration of 20s. This subscription allows you to upload videos without watermarks and create up to five videos at a time.

In both cases, videos can be widescreen, vertical or 4:3 (square) aspect ratio.

OpenAI has also introduced a "storyboard" feature that allows you to create videos from photos, as well as create videos with additional text prompts.

The "remix" option also deserves attention - it will allow you to edit already generated videos and combine two clips into one.

Videos created with Sora are watermarked with special watermarks and contain C2PA metadata that indicates the use of AI in their creation.

Sora is currently running worldwide, except in the EU, UK and China. OpenAI's CEO said that launching the AI in "most of Europe and Britain" could "take some time".

Source: The Verge