VLM Run icon

VLM Run

Interact with VLM Run API

Actions6

Overview

The node integrates with the VLM Run API to analyze various types of media files and execute agents. Specifically, for the Analyze Image operation, it processes image data to extract information or generate captions from images. This is useful in scenarios such as:

  • Automatically generating descriptive captions for images to improve accessibility.
  • Extracting metadata or identifying objects within images for content management.
  • Enhancing workflows that require image understanding, like tagging or classification.

Practical example: You could use this node to analyze product images uploaded to an e-commerce platform, automatically generating alt text or extracting key visual features for search indexing.

Properties

Name Meaning
File The image file data to analyze, typically passed from a previous node (e.g., binary data reference).
Domain Name or ID The domain context to use for analysis. Choose from a predefined list or specify an ID via expression.
Callback URL (Optional) URL to be called when asynchronous processing completes. Required if asynchronous mode is enabled.

Note: For the Analyze Image operation, the "Process Asynchronously" property is not listed in the provided properties, so processing is synchronous by default.

Output

The node outputs a JSON object containing the results of the image analysis. The exact structure depends on the VLM Run API response but generally includes extracted information or generated captions related to the input image.

If the node supports binary data output (not explicitly shown here), it would represent processed image data or related artifacts; however, based on the code, the output is JSON only.

Example output snippet (conceptual):

{
  "caption": "A group of people hiking in a forest",
  "objects": [
    {"name": "person", "confidence": 0.98},
    {"name": "tree", "confidence": 0.95}
  ],
  "metadata": {
    "imageWidth": 1024,
    "imageHeight": 768
  }
}

Dependencies

  • Requires an API key credential for authenticating with the VLM Run API.
  • Depends on the VLM Run service being accessible and configured properly.
  • Uses internal helper functions to process files and send requests to the API.
  • If asynchronous processing is enabled (not shown for image operation), requires a valid callback URL endpoint to receive completion notifications.

Troubleshooting

  • File Not Found or Invalid: Ensure the "File" property references valid image data from a previous node.
  • Domain Loading Failure: If domains fail to load, check network connectivity and API credentials.
  • API Errors: Errors returned from the VLM Run API will be surfaced; verify API limits, authentication, and request parameters.
  • Callback URL Issues: If asynchronous processing were enabled, ensure the callback URL is reachable and correctly handles incoming requests.
  • Unsupported Operation: Selecting an unsupported operation will throw an error indicating the operation is not supported.

Links and References

Discussion