Skip to main content

Create Images From Image

Generates new images based on an input image and text prompt using Stability AI's image-to-image generation.

Common Properties​

  • Name - The custom name of the node.
  • Color - The custom color of the node.
  • Delay Before (sec) - Waits in seconds before executing the node.
  • Delay After (sec) - Waits in seconds after executing node.
  • Continue On Error - Automation will continue regardless of any error. The default value is false.
info

If the ContinueOnError property is true, no error is caught when the project is executed, even if a Catch node is used.

Inputs​

  • Connection Id - The connection ID from the Connect node (optional if API Key is provided directly).
  • Image Save Path - Directory path where generated images will be saved. Images are automatically named as img2img_0.png, img2img_1.png, etc.
  • Prompt - Text description of the desired image modifications or transformations.
  • Init Image Path - Path to the source image file to transform. This is the starting point for the generation.

Options​

Authentication​

  • API Key - Stability AI API key (optional if using Connection ID). You can provide the API key directly instead of using a Connect node.

Model Selection​

  • Engine Id - Stability AI engine to use for image generation:
    • Stable Diffusion XL 1.0 - Latest SDXL model, highest quality 1024x1024 images
    • Stable Diffusion XL 0.9 - Previous SDXL version
    • Stable Diffusion 1.5 - Classic SD 1.5, balanced performance (default)
    • Stable Diffusion 2.1 - SD 2.1 for 512x512 images

Generation Settings​

  • Negative Prompt - Text describing elements to avoid in the generated image. Helps exclude unwanted features.
  • Cfg Scale - How strictly the diffusion process adheres to the prompt (default: 7). Range: 1-35. Higher values follow the prompt more closely.
  • Samples - Number of images to generate (1-10, default: 1). Generates multiple variations from the same input.
  • Seed - Random noise seed (default: 0 for random). Use specific values for reproducible results.
  • Steps - Number of diffusion steps to run (10-150, default: 50). More steps = better quality but slower generation.

Init Image Control​

  • Init Image Mode - Mode for controlling init image influence:

    • IMAGE_STRENGTH - Control influence with a single strength value (default, recommended for most use cases)
    • STEP_SCHEDULE - Fine-grained control over which diffusion steps use the init image

  • Image Strength - (IMAGE_STRENGTH mode only) How much the output resembles the init image (0.0-1.0, default: 0.35):

    • 0.0-0.3 - Significant changes, loosely based on input
    • 0.3-0.5 - Moderate changes, balanced transformation (recommended)
    • 0.5-0.7 - Conservative changes, close to original
    • 0.7-1.0 - Minimal changes, very close to original

  • Step Schedule Start - (STEP_SCHEDULE mode only) Skip proportion of start diffusion steps (0.0-1.0, default: 0.65). Lower values = more init image influence.

  • Step Schedule End - (STEP_SCHEDULE mode only) Skip proportion of end diffusion steps (0.0-1.0, optional). Lower values = more init image influence.

Advanced Settings​

  • Clip Guidance Preset - CLIP guidance preset for image generation quality:

    • NONE - No CLIP guidance (default)
    • FAST_BLUE, FAST_GREEN, SIMPLE, SLOW, SLOWER, SLOWEST

  • Sampler - Sampling algorithm used for diffusion process:

    • NONE - Use engine default (recommended)
    • DDIM, DDPM, K_DPMPP_2M, K_DPMPP_2S_ANCESTRAL, K_DPM_2, K_DPM_2_ANCESTRAL, K_EULER, K_EULER_ANCESTRAL, K_HEUN, K_LMS

  • Style Preset - Style preset to guide the image generation:

    • NONE - No style preset (default)
    • 3d-model, analog-film, anime, cinematic, comic-book, digital-art, enhance, fantasy-art, isometric, line-art, low-poly, modeling-compound, neon-punk, origami, photographic, pixel-art, tile-texture

How It Works​

The Create Images From Image node transforms an existing image based on a text prompt. When executed, the node:

  1. Validates the connection or creates a temporary client using provided credentials
  2. Validates the prompt, init image path, and save path
  3. Reads the init image file from the specified path
  4. Constructs the generation request with:
    • Main prompt (positive weight)
    • Negative prompt if provided (negative weight)
    • Init image data
    • Init image mode and strength/schedule parameters
    • Generation parameters (steps, cfg_scale, seed, etc.)
    • Style and sampler settings
  5. Sends the multipart request to Stability AI API
  6. Receives base64-encoded images from the API
  7. Decodes and saves images to the specified directory
  8. Images are named sequentially: img2img_0.png, img2img_1.png, etc.

Requirements​

  • Either a valid Connection Id from Connect node OR direct API Key credentials
  • Non-empty Prompt
  • Valid Init Image Path (file must exist and be a valid image)
  • Valid Image Save Path (directory must exist)
  • Sufficient Stability AI credits for image generation

Error Handling​

The node will return specific errors in the following cases:

  • Empty or missing Prompt
  • Empty or missing Init Image Path
  • Empty or missing Image Save Path
  • Invalid Connection Id (when not using direct credentials)
  • Init image file not found at specified path
  • Invalid image file format
  • Cfg Scale, Samples, Seed, or Steps not valid integers
  • Image Strength, Step Schedule Start/End not valid decimal numbers
  • Samples outside range of 1-10
  • Steps outside range of 10-150
  • API authentication errors (401)
  • Insufficient credits (402)
  • API rate limit errors (429)
  • API service errors (500, 503)

Usage Notes​

Image-to-Image vs Text-to-Image​

  • Use TextToImages when starting from scratch with just a prompt
  • Use ImageToImages when you want to modify or transform an existing image
  • Image-to-image is ideal for:
    • Style transfer (change art style of existing image)
    • Image variations (create similar but different versions)
    • Enhancement and refinement
    • Guided transformations

Image Strength Guidelines​

  • 0.1-0.2: Almost completely new image, loosely inspired by original
  • 0.3-0.4: Significant transformation while keeping main structure
  • 0.5-0.6: Moderate changes, recognizable as transformation of original
  • 0.7-0.8: Conservative changes, maintains most details
  • 0.9-1.0: Very subtle changes, almost identical to original

Init Image Modes​

  • IMAGE_STRENGTH (recommended): Simple, intuitive control with single parameter
  • STEP_SCHEDULE: Advanced control for specific artistic effects, requires experimentation

Supported Image Formats​

  • PNG, JPEG, JPG, WebP, GIF
  • Recommended: Use PNG or JPEG for best results
  • Image size should match the engine's capabilities

Examples​

Example 1: Style Transfer​

Inputs:

  • Connection Id: (from Connect node)
  • Image Save Path: "/path/to/output"
  • Prompt: "Van Gogh painting style, starry night, impressionist brushstrokes"
  • Init Image Path: "/path/to/photo.jpg"

Options:

  • Engine Id: Stable Diffusion XL 1.0
  • Image Strength: 0.4
  • Steps: 50
  • Style Preset: analog-film

Use Case: Transform a regular photo into Van Gogh style artwork

Example 2: Photo Enhancement​

Inputs:

  • Connection Id: (from Connect node)
  • Image Save Path: "/path/to/output"
  • Prompt: "High quality professional photo, enhanced details, vibrant colors, sharp focus"
  • Init Image Path: "/path/to/original.jpg"
  • Negative Prompt: "blurry, low quality, noise, artifacts"

Options:

  • Engine Id: Stable Diffusion XL 1.0
  • Image Strength: 0.7
  • Steps: 40
  • Style Preset: enhance

Use Case: Enhance and improve photo quality while maintaining original composition

Example 3: Seasonal Transformation​

Inputs:

  • Connection Id: (from Connect node)
  • Image Save Path: "/path/to/output"
  • Prompt: "Winter scene, snow covered, frozen landscape, cold atmosphere"
  • Init Image Path: "/path/to/summer_photo.jpg"

Options:

  • Engine Id: Stable Diffusion 1.5
  • Image Strength: 0.35
  • Steps: 50

Use Case: Transform a summer landscape into a winter scene

Example 4: Multiple Variations​

Inputs:

  • Connection Id: (from Connect node)
  • Image Save Path: "/path/to/output"
  • Prompt: "Fantasy illustration, magical atmosphere, mystical lighting"
  • Init Image Path: "/path/to/base_image.jpg"

Options:

  • Engine Id: Stable Diffusion XL 1.0
  • Image Strength: 0.4
  • Samples: 5
  • Steps: 50
  • Seed: 0

Result: 5 different fantasy variations of the same base image

Example 5: RPA Use Case - Product Image Variations​

Flow:

Read product images from folder
  → Loop through each image
    → ImageToImages (create variations)
    → Rename with product SKU + variant number
    → Upload to e-commerce platform

Inputs:

  • Prompt: "Professional product photography, white background, studio lighting, e-commerce style"
  • Init Image Path: msg.productImagePath
  • Image Save Path: "/temp/product_variants"

Options:

  • Engine Id: Stable Diffusion XL 1.0
  • Image Strength: 0.6
  • Samples: 3
  • Style Preset: photographic

Use Case: Generate multiple professional variants of product photos for A/B testing

Example 6: RPA Use Case - Batch Style Application​

Flow:

Read image list from CSV
  → Get style template from database
  → Loop through images
    → ImageToImages (apply consistent style)
    → Watermark (add company logo)
    → Save to output folder

Inputs:

  • Prompt: msg.styleDescription (e.g., "Corporate professional style, blue color scheme, modern aesthetic")
  • Init Image Path: msg.imagePath
  • Image Save Path: "/output/styled_images"

Options:

  • Engine Id: Stable Diffusion 1.5
  • Image Strength: 0.5
  • Style Preset: photographic

Use Case: Apply consistent branding style to batch of images

Example 7: RPA Use Case - Real Estate Virtual Staging​

Flow:

Read empty room photos
  → ImageToImages (add furniture and decor)
  → Resize for listing platform
  → Upload to real estate website

Inputs:

  • Prompt: "Beautifully furnished living room, modern furniture, warm lighting, interior design magazine quality, cozy atmosphere"
  • Init Image Path: "/photos/empty_room.jpg"
  • Negative Prompt: "empty, bare, cluttered, messy"

Options:

  • Engine Id: Stable Diffusion XL 1.0
  • Image Strength: 0.35
  • Steps: 50
  • Style Preset: photographic

Use Case: Virtually stage empty rooms for real estate listings

Example 8: RPA Use Case - Image Restoration and Modernization​

Flow:

Read archive of old product photos
  → ImageToImages (modernize and enhance)
  → Quality check
  → Update product catalog database

Inputs:

  • Prompt: "Modern professional product photo, high quality, current photography trends, clean background"
  • Init Image Path: msg.oldPhotoPath
  • Negative Prompt: "vintage, dated, low quality, blurry"

Options:

  • Engine Id: Stable Diffusion XL 1.0
  • Image Strength: 0.6
  • Steps: 40
  • Style Preset: enhance

Use Case: Update old product photography to modern standards

Example 9: Anime to Realistic Conversion​

Inputs:

  • Connection Id: (from Connect node)
  • Image Save Path: "/path/to/output"
  • Prompt: "Photorealistic portrait, natural lighting, detailed skin texture, professional photography"
  • Init Image Path: "/path/to/anime_character.jpg"
  • Negative Prompt: "cartoon, anime, illustration, drawn"

Options:

  • Engine Id: Stable Diffusion XL 1.0
  • Image Strength: 0.3
  • Steps: 50
  • Style Preset: photographic

Use Case: Convert anime/cartoon images to photorealistic style

Example 10: Advanced - Step Schedule Control​

Inputs:

  • Connection Id: (from Connect node)
  • Image Save Path: "/path/to/output"
  • Prompt: "Oil painting masterpiece, classical art style"
  • Init Image Path: "/path/to/photo.jpg"

Options:

  • Engine Id: Stable Diffusion 1.5
  • Init Image Mode: STEP_SCHEDULE
  • Step Schedule Start: 0.5
  • Step Schedule End: 0.1
  • Steps: 100

Use Case: Fine-grained control over init image influence for specific artistic effects

Best Practices​

  1. Image Strength Selection:

    • Start with 0.4-0.5 for balanced transformations
    • Lower (0.2-0.3) for dramatic changes
    • Higher (0.6-0.8) to preserve original details
    • Test different values to find optimal balance

  2. Prompt Writing:

    • Describe the desired transformation clearly
    • Include target style, mood, and atmosphere
    • Use negative prompts to avoid unwanted elements
    • Be specific about what should change vs. what should stay

  3. Init Image Quality:

    • Use high-resolution source images when possible
    • Ensure init image is clear and well-lit
    • Clean up or crop images before processing
    • Avoid heavily compressed or artifacted images

  4. Engine Selection:

    • SDXL 1.0: Best quality for realistic transformations
    • SD 1.5: Good balance of quality and speed
    • Match engine to desired output style

  5. Generation Parameters:

    • Use 40-50 steps for production quality
    • Cfg Scale 7-10 for most transformations
    • Generate multiple samples to choose best result
    • Use consistent seeds for reproducible transformations

  6. Cost Optimization:

    • Test with lower steps first (30-40)
    • Generate single sample initially, then batch
    • Use image strength to control processing intensity
    • Monitor credit usage with GetUserCredit

  7. Quality Control:

    • Always use negative prompts
    • Review generated images before batch processing
    • Adjust strength based on initial results
    • Keep successful parameters for future use

Tips for RPA Developers​

  • Batch Processing: Loop through image folders for bulk transformations
  • Error Handling: Implement try-catch for file not found and API errors
  • File Validation: Check init image exists and is valid format before calling
  • Output Organization: Create dated folders or category-based organization
  • Progress Tracking: Log successful transformations and failures
  • Quality Assurance: Implement automated or manual QA step after generation
  • Fallback Strategy: Have default image strength values that work for most cases
  • A/B Testing: Generate variations to compare transformation results