Skip to main content
The Topaz Image Enhance node provides industry-standard upscaling and image enhancement. It processes a single input image using a cloud-based AI model to improve quality, detail, and resolution. The node offers fine-grained control over the enhancement process, including options for creative guidance, subject focus, and facial preservation.

Inputs

ParameterDescriptionData TypeRequiredRange
modelThe AI model to use for image enhancement.COMBOYes"Reimagine"
imageThe input image to be enhanced. Only one image is supported.IMAGEYes-
promptOptional text prompt for creative upscaling guidance (default: empty).STRINGNo-
subject_detectionControls which part of the image the enhancement focuses on (default: “All”).COMBONo"All"
"Foreground"
"Background"
face_enhancementEnhance faces (if present) during processing (default: True).BOOLEANNo-
face_enhancement_creativitySet the creativity level for face enhancement (default: 0.0).FLOATNo0.0 - 1.0
face_enhancement_strengthControls how sharp enhanced faces are relative to the background (default: 1.0).FLOATNo0.0 - 1.0
crop_to_fillBy default, the image is letterboxed when the output aspect ratio differs. Enable to crop the image to fill the output dimensions (default: False).BOOLEANNo-
output_widthZero value means to calculate automatically (usually it will be original size or output_height if specified) (default: 0).INTNo0 - 32000
output_heightZero value means to output in the same height as original or output width (default: 0).INTNo0 - 32000
creativityControls the overall creativity level of the enhancement (default: 3).INTNo1 - 9
face_preservationPreserve subjects’ facial identity (default: True).BOOLEANNo-
color_preservationPreserve the original colors (default: True).BOOLEANNo-
Note: This node can only process a single input image. Providing a batch of multiple images will result in an error.

Outputs

Output NameDescriptionData Type
imageThe enhanced output image.IMAGE
This documentation was AI-generated. If you find any errors or have suggestions for improvement, please feel free to contribute! Edit on GitHub

Source fingerprint (SHA-256): 586ebc6aafa40c8bebf1e24904ebf5653b8d3b46b202d9a6850f625ac8845bcf