TD-XL A.I. Art Model

Learn More - TDXL

TD-XL is a groundbreaking and free AI model we created for the open source community. It’s the premier model for photorealism and all-around performance, all without the steep learning curve of complex prompting.

Just a few of the thousands of images created with TD-XL

PROBLEM

As avid users of stable diffusion, we’ve struggled to find a go-to model that’s great at photorealism and also versatile enough to generate high-quality images across a variety of styles and subjects without needing to be prompting geniuses.

So we set an ambitious target to build our own.

SOLUTION

Designed for artists, creators, and developers, TDXL is not just another AI model; it’s the premier model for photorealism and all-around performance, all without the steep learning curve of complex prompting.

With over 750 hours dedicated to training and meticulous hand-captioning of images taking up to 400 hours, TDXL is a model that stands out in the crowd. The training involved 1.8 million steps, dwarfing other models like Juggernaut and RealVisXL, and was conducted over a 42GB dataset of more than 10,000 diverse images.

With the hurculean training comes unmatched detail and quality, a more balanced and less biased artistic experience, and superior ease of use with simple prompting that yields high-quality images with less effort

Now available on all open source model download sites and preinstalled at http://thinkdiffusion.com