Ask AI
Skip to main content

LLM Image creation request

Function: LLM Image creation request​

Send a prompt to an AI model to generate an image based on your instructions. This action allows you to describe the image you want, choose a specific AI model, and receive the generated image as a result.

Input​

  • Instructions (Required)
    • A clear and detailed description of the image you want the AI to create. Think of it as telling an artist exactly what to paint.
  • Instructions placeholders (Optional)
    • If your main "Instructions" contain special tags (like \{\{COLOR\}\} or \{\{OBJECT\}\}), you can use this input to automatically replace those tags with specific values. This is useful for creating dynamic instructions without having to rewrite the entire prompt each time.
  • Model (Required)
    • Choose the specific AI model that will generate your image. Different models may have unique styles, capabilities, or specialties. You'll select from a list of available models.
  • Files (Optional)
    • You can provide existing image files as a reference or a starting point for the AI. Please note that not all AI models support using reference files. This input accepts a list of files.
  • API Token (optional) (Optional)
    • If you have a specific API key for the AI service, you can enter it here. If you leave this field blank, the platform will use your company's default API key settings.

Output​

  • Result
    • This is the variable where the generated image (or a list of images, if the model creates multiple) will be stored. The output will be a list of files.

Execution Flow​

Real-Life Examples​

Example 1: Generating a Product Image for an E-commerce Store

Imagine you need a new hero image for a product page.

  • Inputs:
    • Instructions: "A sleek, modern smartphone, silver color, on a minimalist white background with soft studio lighting. Show the screen displaying a vibrant app icon."
    • Model: "DALL-E 3" (or a similar image generation model)
    • Result: "SmartphoneHeroImage"
  • Result: The action generates a high-quality image of a silver smartphone as described and stores it in the SmartphoneHeroImage variable, ready to be used on your product page.

Example 2: Creating Variations of a Marketing Banner with Placeholders

You want to create several marketing banners with slight variations for different campaigns.

  • Inputs:
    • Instructions: "A vibrant banner for a \{\{SEASON\}\} sale. Feature a \{\{PRODUCT_TYPE\}\} with a \{\{DISCOUNT\}\} off. Use a playful, energetic style."
    • Instructions placeholders:
      • SEASON: "Summer"
      • PRODUCT_TYPE: "new collection of shoes"
      • DISCOUNT: "20%"
    • Model: "Midjourney v5"
    • Result: "SummerSaleBanner"
  • Result: The action generates a marketing banner for a "Summer sale" featuring a "new collection of shoes" with "20% off" in a playful style, storing the image in SummerSaleBanner. You could then run this action again with different placeholder values (e.g., SEASON: "Winter", PRODUCT_TYPE: "cozy sweaters", DISCOUNT: "30%") to generate new banners.

Example 3: Transforming a Sketch into a Digital Painting

You have a rough sketch of a character and want to turn it into a polished digital painting.

  • Inputs:
    • Instructions: "Transform this sketch into a detailed, fantasy-style digital painting of a knight. Emphasize metallic armor and a glowing sword."
    • Files: [Upload knight_sketch.png]
    • Model: "Stable Diffusion XL" (assuming this model supports image-to-image generation)
    • Result: "FantasyKnightPainting"
    • API Token (optional): "your_custom_api_key_here"
  • Result: The action takes your knight_sketch.png as a reference and generates a detailed digital painting of a fantasy knight, storing the final image in the FantasyKnightPainting variable.