LiblibAI Cutout is an AI background removal workflow from LiblibAI that takes product images out of their original background and places them in a new one: upload a product photo plus a style reference and the system detects the subject, generates a background and adjusts lighting and tone, which suits e-commerce detail pages and posters.
What is LiblibAI Cutout
LiblibAI Cutout is an AI cutout workflow launched by LiblibAI, used mainly for one-click background removal in e-commerce pipelines. You upload a product image and it detects the subject and cuts it out automatically. Then you can choose from several built-in background styles or upload a background reference with a similar style; the system creates a new background from that reference and adjusts light, shadow and color so the product blends in naturally. The whole flow is simple, which makes it a fit for e-commerce beginners and for users without a design background who want product shots that look more professional.
Main capabilities
- Smart detection and cutout: Detects the subject in the product image and cuts it out precisely, giving a clean base for later steps such as background replacement.
- Background replacement and blending: Upload a background reference with a similar style and the system generates a new background from it, adjusting light, shadow and color so the product blends in naturally.
- Easy to run: Both online and local runs are straightforward: online you just open the workflow and upload files; local runs need some setup but suit users who want to use their own hardware performance.
How to use LiblibAI Cutout
- Get the workflow: open the official site www.liblib.art/workflows , go to the workflow selection screen, pick the cutout feature and download the workflow or app.
- Choose how to run it: run it online or locally. Online suits beginners, just open the workflow and upload files; local runs require installing ComfyUI and the related node models and adjusting parameters.
- Upload images: prepare the product image and the reference image and upload both. The product image is the shot to process; the reference image is a background with a similar style.
- Click generate: once the uploads are confirmed, click generate. The workflow cuts out the product first, then builds the background from the reference and adjusts light and color.
- Wait for the result: the whole process takes about 1 to 3 minutes depending on image size and hardware, then you can export the result.
Use cases
- E-commerce product photos: cut products out of their original background and place them in a more appealing one for detail pages and promotional posters.
- Social media content: swap backgrounds so images match a specific theme or campaign and stand out more.
- Design and creative work: designers extract elements and recombine them in new scenes to build concept images or test visual ideas quickly.
- Advertising and marketing assets: blend products or people with campaign backgrounds for more persuasive ad material.
- Personal photo editing: cut out and replace backgrounds in personal photos for digital albums, social sharing or printing.
Limits and notes
- Image quality: the better the product and reference images, the better the result usually is.
- Hardware: for local runs make sure the machine meets the requirements; the graphics card should have at least 8 GB of memory.
- Generation time: it varies with image size and hardware, about 1 to 3 minutes, so allow some waiting time.
- Local runs require installing ComfyUI and the related node models plus configuration, which is a hurdle for beginners.
FAQ
Who is LiblibAI Cutout for?
It suits e-commerce beginners and users without a design background, as well as operations and design staff who process product shots in batches or build posters and ad material.
What is the difference between online and local runs?
Online runs just need the workflow opened and files uploaded, which suits beginners; local runs need ComfyUI and related node models installed plus parameter tuning, but they let users rely on their own hardware, with at least 8 GB of graphics memory.
How long does one generation take?
The whole process takes about 1 to 3 minutes, depending on image size and hardware performance.
Which images do I need to provide?
You need a product image, the shot to be processed, and a reference image, a background with a similar style; the system then creates a new background from the reference and adjusts light and color.
Frequently asked questions
Who is LiblibAI Cutout for?
It suits e-commerce beginners and users without a design background, as well as operations and design staff who process product shots in batches or build posters and ad material.
What is the difference between online and local runs?
Online runs just need the workflow opened and files uploaded, which suits beginners; local runs need ComfyUI and related node models installed plus parameter tuning, but they let users rely on their own hardware, with at least 8 GB of graphics memory.
How long does one generation take?
The whole process takes about 1 to 3 minutes, depending on image size and hardware performance.
Which images do I need to provide?
You need a product image, the shot to be processed, and a reference image, a background with a similar style; the system then creates a new background from the reference and adjusts light and color.
