Flux.1 t2i and AuraSR upscale workflow

Flux, the largest SOTA open-source text-to-image model to date, was developed by Black Forest Labs and is now available on fal. Flux pushes the boundaries of creativity and performance with an impressive 12B, with aesthetics reminiscent of Midjourney.

Dev version: The base model, open sourced under a non-commercial license, for the community to build on.

Schell version: A stripped-down version of the base model, 4-step plotting, 10x faster.

Thanks to the integration of fal's cutting-edge inference engine, the Flux model runs up to 2x faster than eager torch. As a result, processing times are reduced while maintaining superior quality and detail.

Key features：

Enhanced Image Quality:Generate stunning visuals at higher resolutions.

Advanced Human Anatomy and Realism: Achieve highly realistic and anatomically accurate images.

Improved timeliness: Get more accurate and relevant images based on inputs.

Superior speed: Benefit from the speed and efficiency of the Flux Schnell, ideal for demanding applications.

Node installation: Upgrade to the latest Comfyui.

Model Download:

Schnell version

https://huggingface.co/black-forest-labs/FLUX.1-schnell/tree/main

Dev version

https://huggingface.co/camenduru/FLUX.1-dev/tree/main

put t5xxl_fp16.safetensors or t5xxl_fp8_e4m3fn.safetensors and clip_l.safetensors to directory ComfyUI/models/clip/ .

put ae.sft to directory ComfyUI/models/vae/ .

and put flux1-schnell.sft or flux1-dev.sft to directory ComfyUI/models/unet/ . File size 22G!