HunyuanImageEditApi - ComfyUI Built-in Node Documentation

The Tencent HY Image: Edit node edits or combines reference images from a text instruction with Tencent’s Hunyuan Image model.

The Tencent HY Image: Edit node edits or combines reference images from a text instruction with Tencent’s Hunyuan Image model. Connect one to five images, describe the change in the prompt and refer to the images as @Image1, @Image2 and so on. The node uploads the reference images, sends the request to the API, and returns the edited result.

Inputs

Common Inputs

Parameter	Description	Data Type	Required	Range
model	The model used for editing. The selected model determines which additional inputs are shown.	DYNAMIC_COMBO	Yes	"hy-image-3.5-preview"

hy-image-3.5-preview Inputs

Parameter	Description	Data Type	Required	Range
prompt	The editing instructions. Supports @Image1-style references to the connected images. Must not be empty (default: empty).	STRING	Yes	Any text
aspect_ratio	Aspect ratio of the output. "auto" follows the aspect ratio of the first reference image and is not available at 4K. Ignored when resolution is "custom".	COMBO	Yes	"auto" "1:1" "3:2" "2:3" "4:3" "3:4" "16:9" "9:16" "21:9" "9:21" (default: "auto")
resolution	Pixel area of the output: 1K is about 1024x1024, 2K about 2048x2048 and 4K about 4096x4096. Anything above 2K is rendered at 2K and upscaled by the model. Set to "custom" to use width and height instead of a preset area.	COMBO	Yes	"1K" "2K" "4K" "custom" (default: "2K")
width	Width of the output in pixels. Used only when resolution is "custom".	INT	Yes	256-8192, steps of 16 (default: 2048)
height	Height of the output in pixels. Used only when resolution is "custom".	INT	Yes	256-8192, steps of 16 (default: 2048)
seed	Seed used for generation. Results still vary between runs with the same seed.	INT	Yes	0-2147483647 (default: 42)
reference_detail	How much detail of the reference images the model sees: "standard" allows up to 1024x1024 pixels per image, "high" up to 2048x2048 and preserves small text and fine detail better, but takes longer. This is an advanced parameter.	COMBO	No	"standard" "high" (default: "standard")
watermark	Whether to add an AI-generated watermark to the result. This is an advanced parameter.	BOOLEAN	No	true false (default: false)

Reference Inputs

Parameter	Description	Data Type	Required	Range
images	The reference images to edit or combine. Growable slot: connect 1 to 5 images (image_1 through image_5). Refer to them in the prompt as @Image1 … @Image5, numbered in input order; a batched input counts once per image.	IMAGE	Yes	1-5 images

prompt must not be empty, and a prompt reference such as @Image3 raises an error when only 1 or 2 images are connected. At most 5 reference images can be used in total, counting each image of a batch separately. With the "custom" resolution, width and height must both be multiples of 16 and their product must not exceed the 4096 x 4096 pixel area limit (about 16.7 megapixels); any aspect ratio works, although beyond roughly 6:1 the model starts repeating the subject. The "auto" aspect ratio only works within the 2K area limit, so at 4K pick an explicit aspect ratio or use "custom".

Outputs

Output Name	Description	Data Type
IMAGE	The edited image produced from the reference images and the prompt.	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): 46c112347b51a2983521f87bbeb047289515f4073c48ba5d909fd3bae4597633