Image Node

Image nodes are used for AI image generation, image editing, style transfer, and multi-image composition.

Generation Modes

Image nodes automatically switch modes based on inputs:

Mode	Trigger Condition	Description
Text to Image	Only Text connection	Generate new images from prompts
Image to Image	Image + Text connection	Generate variants based on reference image
Multi-Image Fusion	Multiple Image connections	Fuse multiple images to generate result

Basic Usage

1. Create New Image

Method 1: Double-click canvas and select image node Method 2: Click image icon in bottom toolbar

Canvas menu showing how to add an Image node

Select model:

Image node panel with model selector highlighted

Expanded image model dropdown list

2. Function Area Overview

Overall structure:

Annotated layout of the Image node interface

Model-specific functions (resolution / format / grid display / copy / download / upscale):

Model-specific controls for resolution format grid copy download and upscale

Main display area:

Main preview area showing generated image results

Upload and asset library loading:

Upload button and asset library loading area in Image node

3. Image Generation

Text to Image

Text-to-image workflow with prompt-only input

You can also connect a Text node for prompt input:

Image node connected to a Text node for prompt input

Image to Image (single reference)

Image-to-image mode with one reference image connected

Image-to-image generation result based on a single reference

Parameters

Prompt input field:

Prompt input field in the Image node settings

Model selector:

Model selector dropdown in the Image node

Aspect ratio selection:

Aspect ratio options for image generation

Generation count:

Generation count control for number of outputs

History / Enhance prompt / Generate button:

History enhance prompt and generate controls

Special Features

Image to Video: Generate new Video node and auto-connect

Image-to-video action creating and auto-connecting a Video node

Image Drawing: Generate drawing node and connect

Image drawing action creating and connecting a Drawing node

Drawing interface:

Drawing interface opened from the Image node

Effect after save and exit:

Canvas after saving drawing changes and exiting editor

Camera Control: Generate camera control node and connect

Camera control action creating and connecting a camera node

Example effect:

Example output using camera control with Image node

Multi-Image Input

Multi-image reference (up to 8 images):

Multi-image reference inputs showing several connected images

Use @assets / templates as reference:

Using @assets and templates as image references

Image Models

Model Name	Core Features & Advantages	Image-to-Image/Editing	Pricing (Regular/Subscribed)
JiMeng 3.0	Cinematic quality, rich aesthetics, up to 2K output	Not supported	2 credits
JiMeng 4.0	Supports 4K, new local editing	Supported	10 / Free
JiMeng 4.5	Upgraded 4.0, stronger fusion capabilities	Supported	10 / Free
Google Nano Banana	Easy to use, multi-image composition, strong character consistency	Supported	10 / Free
Google Nano Banana Pro	4K quality output, stronger multi-image composition	Supported	10 / Free
Google Imagen 4	High quality, precise instruction understanding	Not supported	6 credits
Flux 2	Fast speed, adjustable parameters	Supported	2 credits
Flux 2 Pro	Professional-grade quality enhancement	Supported	6 credits
Flux Dev	Prototype and style fusion	Supported	4 credits
Flux 1.1 Pro	Stable and efficient output	Not supported	12 credits
Flux Kontext	Original modification, style replacement	Supported	6 credits
Flux Kontext Max	Upgraded Kontext	Supported	6 credits
Ideogram 3.0	Strong text rendering, rich presets	Supported	10 credits
Ideogram Character	Enhanced character consistency	Supported	20 credits
Runway Gen4 Image	Balance of quality and efficiency	Supported	7 credits
Vidu Reference	Strong multi-image reference	Supported	13 credits
OpenAI GPT Image 1	Narrative-driven, strong restoration

MJ parameter example:

Midjourney parameter input example in the prompt

Example effect:

Example image output generated with Midjourney parameters