AI Model

Overview

The AI Model node runs inference on a selected machine learning model to detect objects, classify images, or extract features. It acts as the core processing unit in vision flows by sending images to the model and returning predictions.

Use AI Model when:

• You need to apply a specific trained model to an image stream.
• You want to extract polygons, points, or classification labels from visual data.
• You are building a detection or inspection pipeline based on custom dataset training.

💡 Tip

Ensure your model is downloaded and enabled in the Models section of the platform before attempting to select it in this node's configuration.

The inputs and outputs of this node are dynamic and adjust based on the selected model's specification and the chosen output format.

Inputs

Name	Type	Description
Image	image	The source image to be processed by the AI model.

Outputs

The available outputs depend on the Output format parameter.

Mapped format — all results bundled into a single output:

Name	Type	Description
prediction	prediction	A dictionary of prediction results from the active model, organized by prediction class.

Unpacked format — individual attributes as separate outputs, prefixed by the prediction class name (e.g. Defect_polygons, Defect_probability):

Name	Type	Description
{Class}_heatmap	heatmap	Spatial distribution of detected features.
{Class}_points	list[point]	Detected coordinates from the model.
{Class}_polygons	list[polygon]	Geometric outlines of objects detected by the model.
{Class}_class	string	The label of the detected prediction class.
{Class}_probability	number	Confidence score of the classification (0–1).

💡 Tip

The specific unpacked outputs depend on what the selected model produces.

Parameters

Parameter	Default	Valid values	Tunable
Model name	None	Dropdown	❌
Output format	unpacked	unpacked, mapping	❌

Model name

The specific machine learning model used to process input images. The dropdown is populated from models that have been enabled in the platform.

Output format

Determines how prediction data is exposed to downstream nodes.

• All value types (unpacked) — individual attributes (polygons, probability, class, etc.) appear as separate output pins.
• Only model classes (mapping) — all results are bundled into a single prediction output, organized by class. Use Extract Attributes to access individual values.

Example

Defect detection

A flow receives high-resolution images of circuit boards and needs to identify soldering defects.

Configuration:

• Model name: PCB_Solder_Inspector_v2
• Output format: unpacked

Result: The node outputs a list of polygons outlining the defects and a probability score for each detected issue.

Related links

Reference:

• Dynamic AI Model — switch between multiple models at runtime
• Extract Attributes — extract specific attributes from model predictions
• Polygon Threshold — filter detected objects by size or confidence