Composition Node
The Composition node combines multiple input assets and a text prompt to generate a multi-reference scene. Rather than simple layering, it uses AI to intelligently compose elements from several sources into a cohesive output.
What It Does
You connect multiple assets -- characters, products, backgrounds, logos -- along with a descriptive prompt, and the node generates a new image that incorporates references from all connected inputs. This makes it ideal for creating product-in-scene mockups, character compositions, and branded visuals where you need several reference elements to appear together naturally.
Configuration
| Parameter | Description | Details |
|---|---|---|
| Input Assets | Multiple image inputs used as references for the composition | Connect from Asset Nodes, Image Generator Nodes, or Remove Background Nodes |
| Prompt | Text description of the desired composed scene | Free-form text |
Usage
- Drag a Composition node onto the canvas.
- Connect two or more source nodes to its inputs. These can be Asset Nodes, Image Generator Nodes, Remove Background Nodes, or any other image-producing nodes.
- Write a Prompt describing how the elements should be composed (for example, "a person holding the product in a modern kitchen").
- Execute the node to generate the composed scene.
For the cleanest compositions, run your subject assets through the Remove Background Node first. Providing isolated subjects alongside a clear prompt gives the AI the best chance of producing a natural-looking scene.
Related Nodes
- Asset Node -- source the reference images for the composition.
- Prompt Generator Node -- generate the prompt that guides the composition.
- Remove Background Node -- isolate subjects before feeding them into the composition.
- Image Generator Node -- generate images to use as composition inputs.
- Image Expand Node -- expand the composed output if you need a wider frame.