OpenAI, the company behind ChatGPT, has introduced its new artificial intelligence (AI) model: Sora.

This model marks OpenAI’s first foray into generating videos through artificial intelligence, expanding its repertoire of technological tools which includes the ChatGPT text generator and the DALL-E image generator.

This new AI development is capable of generating short videos from text. What has drawn attention to Sora is the high quality of the clips it can produce.

“Sora can generate complex scenes with multiple characters, specific types of movement, and precise details of the subject and background. The model understands not only what the user has asked for in the prompt, but also how those things exist in the physical world.

“The model has a deep understanding of language, which allows it to interpret prompts accurately and generate compelling characters that express vibrant emotions. Sora can also create multiple shots within a single generated video that accurately maintain the characters and visual style,” explains OpenAI on its website.

What is Sora OpenAI?

Sora is a multimodal language model that can generate realistic videos from textual descriptions.

Users simply need to write a description of the scene they want to see, and Sora brings it to life. The model can create videos up to 60 seconds long, with quality comparable to that of a video produced by a professional team.

How does it work?

It uses a diffusion model to generate videos. This means it starts with a video that appears to be filled with static noise and gradually smooths it out to obtain a polished and coherent version.

“Sora is capable of generating entire videos at once or extending generated videos to make them longer. By giving the model a forecast of many frames at once, we have solved a challenging problem of ensuring that a subject remains the same even when it temporarily goes out of view.

“Similar to GPT models, Sora uses a transformer architecture, unlocking superior scaling performance,” explains the company led by Sam Altman.

How to use Sora Open AI?

Users can prompt Sora to generate videos by simply writing a few sentences, choosing between a photorealistic or animated style, and obtaining surprising results in a matter of minutes.

“In addition to being able to generate a video solely from text instructions, the model is capable of taking an existing still image and generating a video from it, animating the contents of the image with accuracy and attention to small detail. The model can also take an existing video and extend it or fill in missing frames.”

When will Sora be available? What we know about the release date

This model stands out among competitors like Midjourney and Stable Diffusion for its ability to create longer, dynamic, and fluid videos. Unlike competing models, which can seem like a sequence of AI-generated stop-motion images, Sora manages to produce videos that feel completely real and cohesive.

However, OpenAI has acknowledged that Sora still has areas requiring improvement, including its understanding of cause and effect, as well as spatial awareness. These challenges, along with security concerns related to the potential for creating deceptive deepfakes, are critical aspects that OpenAI is still working on.

Sora is still in development and not publicly available. OpenAI has begun offering access to a small group of researchers and creatives to test the tool and provide feedback. Sora is expected to be launched to the general public at a date yet to be determined.