GeminiNanoBanana2V2 - ComfyUI Built-in Node Documentation - ComfyUI

This node generates or edits images by sending a text prompt to Google’s Vertex AI API through Gemini image models. It creates new images from a description or modifies existing images using optional reference images.

Inputs

Common Inputs

Parameter	Description	Data Type	Required	Range
`model`	Selects the Gemini image model to use. The chosen model determines the available resolution options and model-specific inputs.	DYNAMIC_COMBO	Yes	`"Nano Banana 2 (Gemini 3.1 Flash Image)"` `"Nano Banana 2 Lite"`
`prompt`	Text prompt describing the image to generate or the edits to apply. Include any constraints, styles, or details the model should follow. Must not be empty. (default: empty)	STRING	Yes	N/A
`seed`	When the seed is fixed to a specific value, the model makes a best effort to provide the same response for repeated requests. Deterministic output isn’t guaranteed. Also, changing the model or parameter settings, such as the temperature, can cause variations in the response even when you use the same seed value. By default, a random seed value is used. (default: 42)	INT	Yes	0 to 18446744073709551615
`response_modalities`	Determines the response format. IMAGE returns only an image; IMAGE+TEXT returns an image and a text response. (default: IMAGE) Advanced parameter.	COMBO	Yes	`"IMAGE"` `"IMAGE+TEXT"`
`system_prompt`	Foundational instructions that dictate an AI’s behavior. Defaults to a built-in prompt that instructs the model to always produce an image. Advanced parameter.	STRING	No	N/A
`temperature`	Controls randomness in generation. Lower is more focused/deterministic. (default: 1.0) Advanced parameter.	FLOAT	No	0.0 to 2.0 (step 0.01)
`top_p`	Nucleus sampling threshold. Lower is more focused, higher more diverse. (default: 0.95) Advanced parameter.	FLOAT	No	0.0 to 1.0 (step 0.01)

Nano Banana 2 (Gemini 3.1 Flash Image) Inputs

Parameter	Description	Data Type	Required	Range
`aspect_ratio`	If set to ‘auto’, matches your input image’s aspect ratio; if no image is provided, a 16:9 square is usually generated. (default: auto)	COMBO	Yes	`"auto"` `"1:1"` `"2:3"` `"3:2"` `"3:4"` `"4:3"` `"4:5"` `"5:4"` `"9:16"` `"16:9"` `"21:9"` `"1:4"` `"4:1"` `"8:1"` `"1:8"`
`resolution`	Target output resolution.	COMBO	Yes	`"1K"` `"2K"` `"4K"`
`thinking_level`	Selects the thinking level used by the model.	COMBO	Yes	`"MINIMAL"` `"HIGH"`

Nano Banana 2 Lite Inputs

Parameter	Description	Data Type	Required	Range
`aspect_ratio`	If set to ‘auto’, matches your input image’s aspect ratio; if no image is provided, a 16:9 square is usually generated. (default: auto)	COMBO	Yes	`"auto"` `"1:1"` `"2:3"` `"3:2"` `"3:4"` `"4:3"` `"4:5"` `"5:4"` `"9:16"` `"16:9"` `"21:9"` `"1:4"` `"4:1"` `"8:1"` `"1:8"`
`resolution`	Target output resolution.	COMBO	Yes	`"1K"`
`thinking_level`	Selects the thinking level used by the model.	COMBO	Yes	`"MINIMAL"` `"HIGH"`

Reference Inputs

Parameter	Description	Data Type	Required	Range
`images`	Optional reference image(s). Up to 14 images total. Growable slot: connect `image_1` through `image_14`.	IMAGE	No	0 to 14 images
`files`	Optional file(s) to use as context for the model. Accepts inputs from the Gemini Generate Content Input Files node.	GEMINI_INPUT_FILES	No	N/A

Note: A maximum of 14 reference images can be connected to the images input; exceeding this limit raises an error.

Outputs

Output Name	Description	Data Type
`IMAGE`	The generated or edited image.	IMAGE
`STRING`	A text description or caption generated by the model. Empty when no text is returned, such as when `response_modalities` is set to `IMAGE`.	STRING
`thought_image`	First image from the model’s thinking process. Only available with `thinking_level` HIGH and IMAGE+TEXT modality.	IMAGE

Note: The STRING output is empty when response_modalities is set to IMAGE. If the model does not generate an image in this mode, the node raises an error suggesting switching to IMAGE+TEXT to view the model’s reasoning.

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): 347d28aeb46aa91f7515a31c385a3e3f805a1861116a21dd2ef6575ab7fd4f3e