Ideogram V3 - ComfyUI Native Node Documentation
Node for creating top-quality images and text rendering using Ideogram’s latest V3 API
This node connects to the Ideogram V3 API to perform image generation tasks.
Currently, this node supports two image generation modes:
- Text-to-Image Mode - Generate new images from text prompts
- Inpainting Mode - Regenerate specific areas by providing an original image and mask
Text-to-Image Mode
This is the default mode, activated when no image or mask inputs are provided. Simply provide a prompt and the desired parameters:
- Describe the image you want in the prompt field
- Select an appropriate aspect ratio or resolution
- Adjust other parameters like magic prompt, seed, and rendering quality
- Run the node to generate the image
Inpainting Mode
Important Note: This mode requires both image and mask inputs. If only one is provided, the node will throw an error.
- Connect the original image to the
image
input port - Create a mask with the same dimensions as the original image, where white areas represent parts to be regenerated
- Connect the mask to the
mask
input port - Describe what you want to generate in the masked area in the prompt
- Run the node to perform local editing
Parameter Descriptions
Basic Parameters
Parameter | Type | Default | Description |
---|---|---|---|
prompt | string | "" | Text prompt describing the content to generate |
aspect_ratio | combo | ”1:1” | Image aspect ratio (text-to-image mode only) |
resolution | combo | ”Auto” | Image resolution, overrides aspect ratio when set |
magic_prompt_option | combo | ”AUTO” | Magic prompt enhancement: AUTO, ON, or OFF |
seed | int | 0 | Random seed value, 0 for random generation |
num_images | int | 1 | Number of images to generate (1-8) |
rendering_speed | combo | ”BALANCED” | Rendering speed: BALANCED, TURBO, or QUALITY |
Optional Parameters
Parameter | Type | Description |
---|---|---|
image | image | Input image for inpainting mode (must be provided with mask) |
mask | mask | Mask for inpainting, white areas will be replaced (must be provided with image) |
Output
Output | Type | Description |
---|---|---|
IMAGE | image | Generated image |
How It Works
The Ideogram V3 node uses state-of-the-art AI models to process user input, capable of understanding complex design intentions and text layout requirements. It supports two main modes:
- Generation Mode: Creates new images from text prompts
- Edit Mode: Uses original image + mask combination, replacing only the areas specified by the mask
Source Code
[Node Source Code (Updated 2025-05-03)]