This FLUX Kontext LoRA was created by reverentelusarca

Here is their X page: https://x.com/el_mejnun

Here is the HuggingFace link to the LoRA: https://huggingface.co/reverentelusarca/flux-kontext-zoom-out-lora

Here is their HuggingFace page: https://huggingface.co/reverentelusarca

LoRA for zooming out—more reliable than separate models for different zoom levels. It works well across various scenes but has a few quirks: pixelated edges in outpainted areas, distortion in tight human close-ups, and no precise zoom control—results vary, so you’ll need to experiment.

To use it, prompt something like:

“Zoom out while keeping the subject unchanged. Expand the canvas evenly and fill new areas with natural, scene-matching detail.”

You can even put your original image in different scenes:

-“zoom out (the subject is driving a car)”

-“zoom out (the subject is in the jungle)”

-“zoom out (the subject is on a river)”