← sources

Flux 3 Klein in ComfyUI | Generate, Edit & Combine Images (Ep39)

source: reddit (reddit:comfyui)published 2026-10-20 15:02ingested 2026-10-21 02:40

retained text

In this episode, we test FLUX 3 Klein, the new image model from Black Forest Labs, in ComfyUI with 19 free workflows. One model can generate an image, edit it, and combine several images into one.

We start with text to image: photos and illustrations from a prompt, a Z-Image realism pass, a prompt upsampler that writes the long prompt for you, text in images, a layout with boxes, and image to prompt to image.

Then we edit one image: change the background, edit only inside a box, change the words in an image, restyle a photo, change the season or the time of day, restore and colorize old photos, make character sheets, and outpaint an image wider.

In the last part, we combine several images: put a character in a scene, change an outfit, put a logo on a product, make a group photo, and make a product ad.

We also look honestly at where Flux 3 Klein struggles, like hands, faces at the normal size, the AI look, and some fonts, and what helps.

You need ComfyUI 0.41 or newer and Pixaroma 1.4.223 or newer. The FLUX 3 files are on Hugging Face and need a free approval: log in and click Agree. The FLUX 3 Klein license lets you use the images for paid work while your yearly revenue is under 5 million dollars.

Every workflow has its own chapter, so you can jump straight to the one you need. All 19 workflows are free to download.

document fetched 2026-10-21 00:41 · canonical https://reddit.com/r/comfyui/comments/1xauho2/flux_3_klein_in_comfyui_generate_edit_combine