Block: Image Generator Block

The Image Generator block creates high-quality images using OpenAI's image generation models. Reach for it whenever a workflow needs to turn a text prompt into an image, with control over resolution, quality, style, and background.

Overview

Property	Value
Type	`image_generator`
Category	tools
Color	`#C2410C`

When to Use

Generate marketing or social media visuals from a text description
Produce illustrations or concept art on demand inside a workflow
Create product mockups or hero images for landing pages
Generate transparent-background assets (with GPT Image) for compositing
Turn upstream AI-generated prompts into finished images automatically
Chain the output image into a downstream vision block for further processing

Configuration

Model

Dropdown to choose the generation model. Defaults to dall-e-3. The selected model controls which Size, Quality, Style, and Background sub-blocks appear.

Label	Id
DALL-E 3	`dall-e-3`
GPT Image	`gpt-image-1`

Prompt

Required long-input field. Describe the image you want to generate. Supports {{blockName.field}} references to inject dynamic content from upstream blocks.

Size (DALL-E 3)

Shown when Model is dall-e-3. Defaults to 1024x1024.

Label	Id
1024x1024	`1024x1024`
1024x1792	`1024x1792`
1792x1024	`1792x1024`

Size (GPT Image)

Shown when Model is gpt-image-1. Defaults to auto.

Label	Id
Auto	`auto`
1024x1024	`1024x1024`
1536x1024	`1536x1024`
1024x1536	`1024x1536`

Quality

Shown when Model is dall-e-3. Defaults to standard.

Label	Id
Standard	`standard`
HD	`hd`

Style

Shown when Model is dall-e-3. Defaults to vivid.

Label	Id
Vivid	`vivid`
Natural	`natural`

Background

Shown when Model is gpt-image-1. Defaults to auto.

Label	Id
Auto	`auto`
Transparent	`transparent`
Opaque	`opaque`

API Key

Required. Your OpenAI API key used to authorize the generation request. Marked as a password field; use {{OPENAI_API_KEY}} to pull the value from an environment secret.

Inputs & Outputs

Inputs:
- prompt (string) — Image description prompt
- model (string) — Image generation model (dall-e-3 or gpt-image-1)
- size (string) — Image dimensions
- quality (string) — Image quality level (DALL-E 3 only: standard or hd)
- style (string) — Image style (DALL-E 3 only: vivid or natural)
- background (string) — Background type (GPT Image only: auto, transparent, or opaque)
- apiKey (string) — OpenAI API key
Outputs:
- content (string) — Generation response: the image URL returned by OpenAI (DALL-E 3) or the string direct-image when base64 data is returned directly (GPT Image)
- image (string) — Base64-encoded image data (PNG); always populated after the proxy fetch
- metadata (json) — Generation metadata, including the model name used

Tools

Image Generator (openai_image) — Calls POST https://api.openai.com/v1/images/generations with the configured model, prompt, size, and model-specific options. For DALL-E 3 responses, the tool fetches the time-limited CDN URL via a server-side proxy and converts the image to a base64 string. For GPT Image, base64 is returned directly by the API. The tool exposes content (URL or identifier), image (base64 data), and metadata (model name) on success.

YAML Example

image_generator_1:
  type: image_generator
  name: "Image Generator"
  inputs:
    model: "dall-e-3"
    prompt: "{{start.input}}"
    size: "1024x1024"
    quality: "standard"
    style: "vivid"
    apiKey: "{{OPENAI_API_KEY}}"
  connections:
    outgoing:
      - target: next-block-id

Image Generator Block