Skip to main content

​
Overview

The Generate Image block creates images from text descriptions using AI. Use it for visual content creation, mockups, illustrations, or any task requiring image generation. Choose the image model in the block settings. The available settings depend on the model.

​
Configuration

Click the settings bar at the top of the configuration panel to open the model and image settings. The bar shows the selected model and its main settings. Prompt and Reference Images appear below the bar.

​
Model

Select the image model in the Model dropdown. Options:
  • GPT Image 2 (default)
  • Nano Banana Pro
  • Nano Banana 2
GPT Image 2 uses Size, Quality, Image Format, and Background. The Nano Banana models use Aspect Ratio and Resolution.

​
Prompt (Required)

Text description of the image you want to generate. Example:
Use clear, descriptive language. Include details about:
  • Subject matter
  • Style (photorealistic, illustration, abstract, etc.)
  • Composition and framing
  • Lighting and mood
  • Colors and textures
You can reference placeholders from previous steps: {{step_1.output}}

​
Image Size

Select the dimensions of the generated image in Size. Available for: GPT Image 2 Options:
  • Square (1024×1024) (default)
  • Portrait (1024×1536)
  • Landscape (1536×1024)
Choose based on your use case:
  • Square: Social media posts, profile images, thumbnails
  • Landscape: Website headers, banners, presentation slides
  • Portrait: Mobile content, story formats, full-page graphics

​
Image Quality

Controls the detail and refinement of the generated image in Quality. Available for: GPT Image 2 Options:
  • High: Maximum detail and refinement
  • Medium: Good balance for most use cases (default)
  • Low: Faster generation, lower detail
Higher quality increases generation time and credit usage.

​
Image Format

File format for the generated image. Available for: GPT Image 2 Options:
  • PNG: Lossless compression with transparency support (default)
  • JPG: Smaller file sizes, best for photographs
  • WebP: Modern format with smaller files and good quality
Choose based on your needs:
  • PNG: Use when quality is critical
  • JPG: Best for photographs and when smaller file sizes matter
  • WebP: Modern browsers, balance of quality and size

​
Aspect Ratio

Select the shape of the generated image. Available for: Nano Banana Pro and Nano Banana 2 Options: 1:1 (Square) (default), 2:3 (Portrait), 3:2 (Landscape), 3:4 (Portrait), 4:3 (Landscape), 4:5 (Portrait), 5:4 (Landscape), 9:16 (Vertical), 16:9 (Widescreen), 21:9 (Ultrawide)

​
Resolution

Select the output resolution. Available for: Nano Banana Pro and Nano Banana 2 Options: 1K, 2K (default), 4K

​
Number of Images

Specify how many variations to generate from the same prompt. Find Number of Images under Advanced Settings. Range: 1–10 images for GPT Image 2. 1–5 images for the Nano Banana models. Default: 1 Generating multiple images lets you choose the best result or create variations. Each image counts toward credit usage.

​
Background

Control the background treatment of the generated image. Find Background under Advanced Settings. Available for: GPT Image 2 Options:
  • Auto (default): Model decides background based on prompt
  • Transparent: Requests a transparent background
  • Opaque: Solid, non-transparent background
Slate currently processes Transparent as Auto, so the image may not have a transparent background.

​
Reference Images

Upload reference images to guide the style, composition, or content of the generated image. Optional field Click Upload to add up to 10 PNG or JPG images, up to 20 MB each. When you provide reference images, the model uses them as visual context:
  • Single reference image: Image-to-image generation with your prompt
  • Multiple reference images: Uses all images for style and context guidance
Use cases:
  • Match a specific visual style
  • Recreate composition with different content
  • Generate variations of an existing image
  • Apply brand visual guidelines
Reference images are processed and used alongside your text prompt to create the final output.

​
Error Handling

Define what happens if the Generate Image block fails. Options:
  1. Terminate workflow: Stop execution immediately
  2. Continue execution: Proceed to next step despite error
  3. Retry: Run the block again until it succeeds, up to the Number of retries you choose (1–5). If every retry fails, the workflow stops at this block.
Select based on whether subsequent steps require the generated image.

​
Best Practices

  • Write detailed, specific prompts for better results
  • Include style descriptors (photorealistic, illustration, watercolor, etc.)
  • Specify lighting, composition, and mood in your prompt
  • Use landscape format for website headers and banners
  • Use portrait format for mobile-first content
  • Generate multiple variations when exploring creative options
  • Reference previous step outputs using {{step_n.output}} syntax
  • Use reference images to maintain consistent visual style across generations
  • Test different quality settings to balance speed and detail needs
  • Remember that higher quality and multiple images increase credit usage

​
Common Use Cases

​
What’s Next

Now that you understand the Generate Image block: