Flux: A new open-source AI image generator outperformed Midjourney, SD3, and Auraflow. 😉

Introducing Flux: The Game-Changing Text-to-Image Model

If you’re into the crypto world and love exploring new technologies, you’re in for a treat. Black Forest Labs, the brilliant minds behind Stable Diffusion, have unleashed Flux—the most extensive open-source text-to-image model on the market, boasting a whopping 12 billion parameters. This cutting-edge model is poised to revolutionize the way we perceive visual content generated by AI, rivaling and potentially outperforming the likes of Midjourney and other existing models, be it open or closed source.

The Three Variations of Flux

Black Forest Labs has rolled out Flux in three distinct iterations, each catering to various needs and preferences:

Flux Dev: An open-source model with a non-commercial license for community development.
Flux Schnell: A faster, distilled version operating up to ten times quicker, available under an Apache 2 license.
Flux Pro: A closed-source version accessible via an API for those seeking top-of-the-line performance.

Availability of Flux Dev and Flux Schnell

Flux Dev and Flux Schnell are readily downloadable on Hugging Face for your convenience.
ComfyUI has received updates to support the new Flux models, enabling seamless integration into local workflows.

Black Forest Labs’ Legacy of Innovation

On a mission to push the boundaries of generative AI for media, Black Forest Labs, the masterminds behind Flux, have a track record of groundbreaking advancements in the AI landscape, including the development of VQGAN, Latent Diffusion, Stable Diffusion models, and Adversarial Diffusion Distillation.

Flux: Setting New Standards in Image Synthesis

According to benchmarking tests, Flux’s models have raised the bar in image synthesis, surpassing established models like Midjourney v6.0, Dall-E 3 (HD), and SD3 Ultra in various aspects such as visual quality, size/aspect variability, and output diversity. Black Forest’s claims about the superiority of its Pro and Dev models signal a significant leap forward in the realm of image generation.