Luma Image to Image - ComfyUI Built-in Node Documentation
Node for modifying images using Luma AI
The Luma Image to Image node allows you to modify existing images using Luma AI technology based on text prompts, while preserving certain features and structure of the original image.
Node Function
This node connects to Luma AI’s text-to-image API, enabling users to generate images through detailed text prompts. Luma AI is known for its excellent realism and detail, particularly excelling at generating photorealistic content and artistic style images.
Parameters
Basic Parameters
Parameter | Type | Default | Description |
---|---|---|---|
prompt | string | "" | Text prompt describing the content to generate |
model | select | - | Select which generation model to use |
aspect_ratio | select | 16:9 | Set the aspect ratio of the output image |
seed | integer | 0 | Seed value to determine if node should rerun, but actual results don’t depend on seed |
style_image_weight | float | 1.0 | Weight of the style image, range 0.02-1.0, only effective when style_image is provided |
Optional Parameters
Without the following parameter inputs, the node functions in text-to-image mode
Parameter | Type | Description |
---|---|---|
image_luma_ref | LUMA_REF | Luma reference node connection, influences generation results through input images, can consider up to 4 images |
style_image | image | Style reference image, uses only 1 image, influences the style of generated images, adjusted through style_image_weight |
character_image | image | Adds character features to the generated results, can be a batch of multiple images, up to 4 images |
Output
Output | Type | Description |
---|---|---|
IMAGE | image | The generated image |
Usage Examples
How It Works
The Luma Image to Image node analyzes the input image and combines it with text prompts to guide the modification process. It uses Luma AI’s generation models to make creative changes to images based on prompts.
Node process:
- First uploads the input image to ComfyAPI
- Then sends the image URL with the prompt to Luma API
- Waits for Luma AI to complete processing
- Downloads and returns the generated image
The image_weight parameter controls the degree of influence from the original image - values closer to 0 will preserve more of the original image features, while values closer to 1 allow for more substantial modifications.
Source Code
[Node Source Code (Updated on 2025-05-05)]