Back to the catalogue
VisionIngestion
Image understanding
Describe, compare or answer questions about an image — attached to a chat turn, or handled by a vision step.
A screenshot of an error, a photo of a delivery note, a chart somebody pasted into a ticket: all of it is evidence, and none of it is text.
Ragenta routes an image to a model that can actually see. A model that has not declared vision support is not offered for the step, so a picture never quietly reaches a model that will answer about it without having looked.