Krea 2 BBOX puts comic panels, characters and exact text where you draw the boxes
Krea 2 BBOX is a free community fine-tune of the open Krea 2 image model that follows a layout you write as boxes: panels, people, objects and word-for-word text each get their own position. It is built for multi-panel comics and storyboard pages, with speech bubbles that stay in their panel and characters that stay the same across panels.
What it does
About 90% of training used a plain-text layout format: one element per line, each with a box of four numbers on a 0 to 1000 grid, x first. Prefixes mark a panel (p), object (o), verbatim text (t) and a person, anime or furry character (pe, ac, fc), while an @ line carries style tags and a ~ line the scene. Ordinary prose prompts still work. It comes in a Turbo version (8 steps, CFG off) and a Raw version (50 steps, CFG 5.5), each with two smoothed EMA copies.
Worth knowing
The model card says writing the coordinates y first is the most common mistake and fails silently: the picture still looks fine, just not where you asked. In ComfyUI the single files go in models/diffusion_models and the stock Krea 2 text encoder, VAE and templates work unchanged. LoRA versions (rank 64 to 256) exist for setups that can only stack adapters, but the author measures them at only about 12 to 38% of the full fine-tune's effect, so use the full 26 GB weights for real layout control. Final weights are epoch 20, published on 27 October; the Krea 2 community licence applies.
Try it yourself
A four-panel comic in which an anime handywoman and an anthropomorphic tabby cat assemble a flat-pack shelf and fail completely. @clean line art, flat colors, light cel shading; Digital illustration p[0,0,500,500] Top-left panel: the handywoman holding up the instruction sheet, confident. p[500,0,1000,500] Top-right panel: the cat batting a single screw off the floor. p[0,500,500,1000] Bottom-left panel: a lopsided half-built shelf wobbling. p[500,500,1000,1000] Bottom-right panel: the shelf collapsed, the cat sitting smugly on the wreckage. t[80,40,420,120]"SOME ASSEMBLY REQUIRED." black text in a rounded speech bubble