Stability's new AI model turns images into 3D sculptures

The Stable Fast 3D can generate detailed 3D content from a single image in just half a second

Tech Desk - Aug 05, 2024

An undated image of the Stable Fast 3D. — Stability AI

AI (Artificial intelligence) has already developed the fascinating ability to generate videos and images based on text prompts, which is a reason to be excited about such technology.

However, a new AI model has demonstrated an even greater ability, Stability AI can generate 3D content via a single image.

The Stable Fast 3D AI model manages to generate a 3D image in half a second, that’s a major leap from its predecessor: the SVD 3D took 10 minutes for a single render. Theoretically, the new model should be able to produce more than 1,000 in that time.

Stability's new AI model bases its generation based on a single image, producing a colourised and texturised 3D render. This technology is bound to be a treat for designers, whether that be for video game developers or architects.

The ability to have a visually specific render of a project that’s being worked on, for instance, an architecture plan, is a revolutionary prospect that is bound to save tons of time and energy.

It’s all part of a wider race taking place between AI heavyweights like, OpenAI and Adobe who’ve come out with their own models with similar capabilities, OpenAI’s POINT=E can make 3D models from text prompts and Adobe’s been experimenting with a few things it could include in FireFly.