Standard search only looks for exact filenames (e.g., img_001.png). Semantic Search is different: it understands the content of your image.

Using Florence-2 (Vision) and Nomic (Embeddings), the system scans your images and creates a mathematical map of their meaning.

This allows you to search for concepts like "a cyberpunk city in the rain" or "red flowers on a table", even if those words are not in the filename. We are thrilled to bring this advanced technology to your local machine!