qwen-edit-apps/qwen-image-edit-plus-lora-fusion

OfficialView on Replicate →

Fusion – Product/object blending that fixes perspective and lighting so the subject melts into a new background via the Fusion LoRA.

Capabilities

Reference ImagesSeed

Cost

Community model (estimated from hardware time)

Input Parameters

Name	Type	Description	Default	Constraints
`image`*	string(uri)	Image file uploaded to Cog (jpeg, png, gif, or webp).	`—`	—
`aspect_ratio`	string	Aspect ratio for the generated image	`"match_input_image"`	match_input_image1:116:99:164:33:4
`disable_safety_checker`	boolean	Disable safety checker for generated images.	`false`	—
`go_fast`	boolean	If num_inference_steps is omitted, true runs the 8-step Lightning preset and false runs the 40-step detailed preset.	`—`	—
`lora_scale`	number	Strength applied to the selected LoRA. Leave blank for the proxy default.	`—`	min: 0, max: 4
`lora_weights`	string	LoRA weights to apply. Leave blank to use the proxy's default adapter.	`""`	—
`num_inference_steps`	integer	Explicit denoising step count (1-40). Leave blank to use the go_fast presets (fast = 8 steps, slow = 40 steps).	`—`	min: 1, max: 40
`output_format`	string	Format of the output images.	`"webp"`	webpjpgpng
`output_quality`	integer	Quality when saving the output images, from 0 to 100. 100 is best quality, 0 is lowest quality. Not relevant for .png outputs.	`95`	min: 0, max: 100
`prompt`	string	Prompt hint: paste the Chinese trigger `溶图,纠正产品透视角度和光影并使产品融入背景` (it means “blend the product into the background with corrected perspective and lighting”), then describe in English the scene and lighting you want.	`"溶图,纠正产品透视角度和光影并使产品融入背景, 描述新的环境与灯光让主体自然融入. Blend the product into a modern showroom with warm diffused lighting."`	—
`seed`	integer	Random seed. Set for reproducible generation.	`—`	—
`true_guidance_scale`	number	True classifier-free guidance scale passed to the pipeline.	`1`	min: 0, max: 10

imagerequiredstring

Image file uploaded to Cog (jpeg, png, gif, or webp).

aspect_ratiostring

Aspect ratio for the generated image

Default: "match_input_image"

match_input_image1:116:99:164:33:4

disable_safety_checkerboolean

Disable safety checker for generated images.

Default: false

go_fastboolean

If num_inference_steps is omitted, true runs the 8-step Lightning preset and false runs the 40-step detailed preset.

lora_scalenumber

Strength applied to the selected LoRA. Leave blank for the proxy default.

min: 0, max: 4

lora_weightsstring

LoRA weights to apply. Leave blank to use the proxy's default adapter.

Default: ""

num_inference_stepsinteger

Explicit denoising step count (1-40). Leave blank to use the go_fast presets (fast = 8 steps, slow = 40 steps).

min: 1, max: 40

output_formatstring

Format of the output images.

Default: "webp"

webpjpgpng

output_qualityinteger

Quality when saving the output images, from 0 to 100. 100 is best quality, 0 is lowest quality. Not relevant for .png outputs.

Default: 95min: 0, max: 100

promptstring

Prompt hint: paste the Chinese trigger `溶图,纠正产品透视角度和光影并使产品融入背景` (it means “blend the product into the background with corrected perspective and lighting”), then describe in English the scene and lighting you want.

Default:

"溶图,纠正产品透视角度和光影并使产品融入背景, 描述新的环境与灯光让主体自然融入. Blend the product into a modern showroom with warm diffused lighting."

seedinteger

Random seed. Set for reproducible generation.

true_guidance_scalenumber

True classifier-free guidance scale passed to the pipeline.

Default: 1min: 0, max: 10

Version: 0fdb72e0360dUpdated: 7/25/20261.8K runs