Prompt-Based Object Detection with Gemini 2.0

This n8n template demonstrates Gemini 2.0's bounding box detection feature, letting you find and highlight objects in any image using plain-text instructions like put a bounding box around all adults with children or mark cars parked outside their spaces. It downloads an image, sends it to Gemini for analysis, and draws the detected bounding boxes directly onto the picture. Useful for developers exploring computer vision use cases such as contextual image search, content moderation, or automated compliance checks, this template shows how to build flexible, natural-language-driven object detection without training a custom model.

Category: Image & Video Generation