This LoRA removes the background from images and isolates the main subject. Simply provide an image, and the model will extract the subject with a clean background removal.
┌─────────────────────────────────────┐
│ ░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ │
│ ░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ │ Background
│ ░░░░░░░░░ [Subject] ░░░░░░░░░░░░░ │ gets removed
│ ░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ │ cleanly
│ ░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ │
└─────────────────────────────────────┘Perfect for product photography, subject isolation, and creative compositions.
| Input | Output (background removed) |
|---|---|
![]() | ![]() |
![]() | ![]() |
![]() | ![]() |
![]() | ![]() |
Remove the background from the imageimport fal_client
def on_queue_update(update):
if isinstance(update, fal_client.InProgress):
for log in update.logs:
print(log["message"])
result = fal_client.subscribe(
"fal-ai/flux-2/klein/4b/base/edit/lora",
arguments={
"prompt": "Remove the background from the image",
"model_name": None,
"loras": [{
"path": "https://huggingface.co/fal/flux-2-klein-4B-background-remove-lora/resolve/main/flux-background-remove-lora.safetensors",
"scale": 1.1
}],
"embeddings": [],
"image_urls": ["https://your-image.png"]
},
with_logs=True,
on_queue_update=on_queue_update,
)
print(result)| File | Use Case |
|---|---|
flux-background-remove-lora.safetensors | fal.ai |
sQn8ANj2lbtrfeemXQta0_pytorch_lora_weights_comfy_converted.safetensors | ComfyUI |