AI Image Generator

Overview

The AI Image Generator node allows your workflow to dynamically generate AI-created images based on a text prompt using OpenAI’s DALL-E models. It supports stylistic presets, adjustable output sizes, quality levels, and optional local downloads of generated images.

This node is particularly useful for:

• Automatically generating images for reports, dashboards, and documents.
• Creating marketing or presentation visuals.
• Producing technical or schematic diagrams from structured prompts.
• Generating preview renders or concept art within automated workflows.

---

Node Type

Property	Value
Category	AI → Image Generation
Task Type	AIImageGenerator
Node Type	Task
Inputs	Standard input data from previous node (optional)
Outputs	Generated image URLs and/or downloaded image file paths

---

Configuration Fields

The node presents a modal configuration form with the following fields:

Label	Type	Required	Description
Prompt	Text	✅ Yes	The main text instruction describing what image to generate. Example: “A solar-powered city skyline at sunset.”
Style Preset	Picklist	Optional	Choose a predefined style for the generated image (see presets below). This appends a stylistic instruction to the prompt.
Model	Picklist	Optional	Choose between dall-e-2 or dall-e-3. Defaults to dall-e-3. Note: DALL-E 3 only supports one image per generation.
Size	Picklist	Optional	Image resolution. Valid options depend on model: • dall-e-2: 256×256, 512×512, 1024×1024 • dall-e-3: 1024×1024, 1024×1792, 1792×1024
Quality	Picklist	Optional	Image detail level. Options: standard (default), hd (where supported).
Image Count	Number	Optional	Number of images to generate (max 1 for DALL-E 3). Default = 1.
Download Images	Toggle	Optional	If enabled, generated images are downloaded and saved into the workflow’s working directory.
ServiceAccount-OpenAI	Service Account Selector	✅ Yes	The OpenAI service account to use for API access. Must be configured in Console → Service Accounts prior to use.

---

Style Presets

The following style presets are available to help guide the AI toward specific aesthetics:

Preset Key	Description
marketing_infographic	Professional infographic with clean typography, charts, and corporate colors.
realistic_render	Photorealistic render with detailed lighting, textures, and depth of field.
technical_diagram	Engineering schematic or technical drawing with labels and annotations.
artistic_illustration	Expressive, colorful, and creative artistic style.
minimalist_design	Clean, simple geometric style with minimal color palette.
vintage_poster	Retro poster with nostalgic design, aged textures, and classic typography.
modern_ui	Sleek, digital user interface concept suitable for apps or websites.
product_mockup	Commercial mockup style with professional studio lighting and shadows.

When a style is selected, its descriptive instruction is prepended to the user’s prompt.

---

Example Usage

Example 1 — Generate a Marketing Infographic

Prompt:

“A visual summary of renewable energy sources with icons for solar, wind, and hydro.”

Style Preset: marketing_infographicModel: dall-e-3Size: 1024x1024Download Images: On

The node will generate a professional infographic and optionally download it to the working folder.

---

Example 2 — Generate a Technical Diagram

Prompt:

“An electrical circuit diagram showing power distribution from a transformer to industrial equipment.”

Style Preset: technical_diagramModel: dall-e-2Image Count: 2

The node generates two schematic-style images, accessible from the output variable Result.

---

Output

The node stores the generated results in a variable called Result, which contains an array of image objects:

[
  {
    "Index": 1,
    "Url": "https://openai-generated-image-url.png",
    "RevisedPrompt": "Final prompt used by the AI",
    "LocalPath": "C:\\Workflows\\run-1234\\ai_generated_image_20251007_120501_1.png"
  }
]

If Download Images is disabled, only the Url and RevisedPrompt fields will be populated.

---

Workflow Integration Notes

• Requires an active OpenAI service account configured under Console → Service Accounts with appropriate API key access.
• Works seamlessly with downstream nodes (e.g. AddFileAttachment, SendEmail, UploadToSharePoint) using the generated file paths.
• Images are saved to the workflow’s workingFolder if downloads are enabled.
• When using DALL-E 3, only one image per execution is permitted.

---

Error Handling

The node will fail gracefully if:

• Required parameters (e.g. Prompt or ServiceAccount-OpenAI) are missing.
• The OpenAI service account is invalid or misconfigured.
• An unsupported size or model combination is used.
• API request fails or no image data is returned.

Errors are logged in the Workflow Execution Log under the AIImageGenerator node.

---

Example Output Log

[INFO] Starting AI Image Generation
[INFO] Retrieving OpenAI service account
[INFO] Generating 1 image(s) with model 'dall-e-3', size '1024x1024', quality 'standard'
[DEBUG] Final prompt: Create a professional marketing infographic style image...
[INFO] Successfully generated 1 image(s)
[INFO] AI Image Generation completed successfully

---

Best Practices

• Use concise, descriptive prompts. Example: “A futuristic office building with solar glass facade, isometric view.”
• Select a Style Preset to guide composition and tone.
• For high quality results, prefer DALL-E 3 with 1024x1024 resolution.
• Enable Download Images if you need local file access for downstream processing.
• If generating multiple images, use DALL-E 2.

---

Known Limitations / To Be Confirmed ⚠️

• [ ] Verify if “Quality: hd” is supported for dall-e-3.
• [ ] Confirm maximum image count for dall-e-2 (assumed up to 10).
• [ ] Confirm if workflow supports Base64 image returns for environments without internet access.

---

Node Summary

Setting	Description
Node Name	AI Image Generator
Task Type	AIImageGenerator
Purpose	Generates images using OpenAI’s DALL-E API from a text prompt
Dependencies	OpenAI Service Account
Output Variable	Result (array of image objects)