MagnificImageUpscalerPreciseV2Node - ComfyUI Built-in Node Documentation

Complete documentation for the MagnificImageUpscalerPreciseV2Node node in ComfyUI. Learn its inputs, outputs, parameters and usage.

The Magnific Image Upscale (Precise V2) node performs high-fidelity image upscaling with precise control over sharpness, grain, and detail enhancement. It processes images through an external API, supporting up to a maximum output resolution of 10060×10060 pixels. The node offers different processing styles and can automatically downscale the input if the requested output would exceed the maximum allowed size.

Inputs

Parameter	Description	Data Type	Required	Range
image	The input image to be upscaled. Exactly one image is required. Minimum dimensions are 160x160 pixels. The aspect ratio must be between 1:3 and 3:1.	IMAGE	Yes	-
scale_factor	The desired upscaling multiplier.	STRING	Yes	"2x" "4x" "8x" "16x"
flavor	Processing style: sublime for general use, photo for photographs, photo_denoiser for noisy photos.	STRING	Yes	"sublime" "photo" "photo_denoiser"
sharpen	Image sharpness intensity. Higher values increase edge definition and clarity. Default: 7.	INT	No	0 to 100
smart_grain	Intelligent grain/texture enhancement to prevent the image from looking too smooth or artificial. Default: 7.	INT	No	0 to 100
ultra_detail	Controls fine detail, textures, and micro-details added during upscaling. Default: 30.	INT	No	0 to 100
auto_downscale	Automatically downscale input image if output would exceed maximum resolution. Default: False.	BOOLEAN	No	-

Note: If auto_downscale is disabled and the requested output size (input dimensions × scale_factor) exceeds 10060x10060 pixels, the node will raise an error. When auto_downscale is enabled, the node will attempt to find an optimal scale factor that keeps quality loss minimal.

Outputs

Output Name	Description	Data Type
image	The resulting upscaled 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): b4de8780c46d113fd043930daa6d4b91df3dca493afef486a0ecf2ee9e0ae708