PiAPI GPT-4o Image to Image

Overview

This node transforms images using the PiAPI GPT-4o Image Generation API. It takes an input image (either from a URL or binary data) and applies a text prompt to generate a transformed version of that image. The node supports different models, multiple input images, various aspect ratios, and artistic styles to customize the output.

Common scenarios include:

• Enhancing or modifying existing images based on descriptive prompts.
• Creating stylized versions of photos for social media or marketing.
• Generating variations of an image with different artistic influences.
• Automating image transformation workflows in creative projects.

For example, you could input a photo URL and a prompt like "turn this into a cyberpunk cityscape" and receive a transformed image styled accordingly.

Properties

Output

The node outputs JSON objects with the following structure:

• prompt: The original text prompt used for transformation.
• status: Status of the generation, either "completed" or "failed".
• If successful:
  • source_image: The URL of the original source image.
  • image_url: The URL of the transformed/generated image (may be null if unavailable).
  • processed_content: Processed content extracted from the API response (likely includes metadata or additional info).
  • original_response: The full raw response from the API.
• If failed:
  • error: An object containing:
    • reason: Reason for failure.
    • suggestion: Suggested fix or workaround.
    • full_message: Full error message or response content.
  • original_response: The full raw response from the API.

If the node receives binary input, it converts it to a base64 data URL format before sending it to the API.

Dependencies

• Requires an API key credential for the PiAPI GPT-4o Image Generation API.
• The node makes HTTP POST requests to https://api.piapi.ai/v1/chat/completions with authentication.
• Uses helper functions for streaming response processing and error extraction.
• Requires n8n environment configured with the appropriate API key credential.

Troubleshooting

• Error: "The provided binary data is not an image"
  Occurs if the binary input does not have an image MIME type. Ensure the binary data contains valid image data.

• Error: "No usable image data found in the provided binary property"
  Happens if the specified binary property lacks usable data or URL. Verify the binary property name and that the data exists.

• API errors or failures
  The node captures failure reasons and suggestions from the API response. Check the error field in the output for details and adjust prompts or inputs accordingly.

• Multiple images input only supported by GPT-4o Image model
  Using multiple images with other models will not work; ensure the correct model is selected.

• Custom aspect ratio requires valid width and height
  When selecting custom aspect ratio, provide sensible numeric values for width and height.

Links and References

• PiAPI GPT-4o Image Generation API Documentation (hypothetical link)
• n8n documentation on Creating Custom Nodes
• General info on Image-to-Image AI Models