OpenAIGPTImageNodeV2 - ComfyUI Built-in Node Documentation

This node generates images using OpenAI’s GPT Image API.

This node generates images using OpenAI’s GPT Image API. It supports multiple models (gpt-image-2, gpt-image-1.5, and gpt-image-1), allows you to provide reference images for editing, and can use a mask to specify which parts of an image to modify.

Inputs

Common Inputs

Parameter	Description	Data Type	Required	Range
model	The OpenAI GPT Image model to use. Selecting a model reveals additional parameters specific to that model.	DYNAMIC_COMBO	Yes	"gpt-image-2" "gpt-image-1.5" "gpt-image-1"
prompt	Text prompt for GPT Image (default: "").	STRING	Yes	N/A
n	How many images to generate (default: 1).	INT	Yes	1 to 8
seed	Seed for reproducibility (default: 0). Not implemented yet in backend.	INT	Yes	0 to 2147483647

gpt-image-2 Inputs

These inputs appear when model is set to gpt-image-2.

Parameter	Description	Data Type	Required	Range
model.size	Image size. Select “Custom” to use the custom width and height (default: "auto").	COMBO	Yes	"auto" "1024x1024" "1024x1536" "1536x1024" "2048x2048" "2048x1152" "1152x2048" "3840x2160" "2160x3840" "Custom"
model.custom_width	Used only when model.size is “Custom”. Must be a multiple of 16 (default: 1024).	INT	No	1024 to 3840
model.custom_height	Used only when model.size is “Custom”. Must be a multiple of 16 (default: 1024).	INT	No	1024 to 3840
model.background	Return image with or without background (default: "auto").	COMBO	Yes	"auto" "opaque"
model.quality	Image quality, affects cost and generation time (default: "low").	COMBO	Yes	"low" "medium" "high"
model.images	Optional reference image(s) for image editing. Up to 16 images. See Reference Inputs for details.	IMAGE	No	0 to 16
model.mask	Optional mask for inpainting (white areas will be replaced). Requires exactly one reference image.	MASK	No	N/A

gpt-image-1.5 and gpt-image-1 Inputs

These inputs appear when model is set to gpt-image-1.5 or gpt-image-1. Both models share the same parameter set.

Parameter	Description	Data Type	Required	Range
model.size	Image size (default: "auto").	COMBO	Yes	"auto" "1024x1024" "1024x1536" "1536x1024"
model.background	Return image with or without background (default: "auto").	COMBO	Yes	"auto" "opaque" "transparent"
model.quality	Image quality, affects cost and generation time (default: "low").	COMBO	Yes	"low" "medium" "high"
model.images	Optional reference image(s) for image editing. Up to 16 images. See Reference Inputs for details.	IMAGE	No	0 to 16
model.mask	Optional mask for inpainting (white areas will be replaced). Requires exactly one reference image.	MASK	No	N/A

Reference Inputs

Parameter	Description	Data Type	Required	Range
model.images	Growable slot: connect 1..N items (e.g. image_1…image_16); up to 16 reference images for all models.	IMAGE	No	1 to 16
model.mask	Optional mask for inpainting (white areas will be replaced). Requires exactly one reference image.	MASK	No	N/A

Parameter Constraints and Limitations:

• When model.size is “Custom” (gpt-image-2 only), model.custom_width and model.custom_height must both be multiples of 16, the longest edge must not exceed 3840, the aspect ratio must not exceed 3:1, and the total pixel count must be between 655,360 and 8,294,400.
• model.mask requires exactly one reference image in model.images: it cannot be used without an image, and it cannot be used with more than one image.
• When model.mask is used, its dimensions must match the dimensions of the reference image.
• When model.images is provided, the node runs in image editing mode; without model.images, it generates images from the prompt alone.
• Reference images are downscaled before being sent to the API.
• seed is currently not implemented in the backend.

Outputs

Output Name	Description	Data Type
image	The generated image or images. All returned images are stacked into one batch; if their dimensions differ, they are resized to match the first 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): fb3491f949151fbd3f5825ec9f9ae124019767d083f56966ef34af278aef50c0