LensLab

Studio

Drop an image and LensLab tells you what it is — no labels needed. It scores the image against a built-in vocabulary and reports kind, subject, setting and look. Want a vocabulary of your own instead? Type your labels below and score against those. Press Ctrl+V anywhere to analyse what's on your clipboard.

1 · The image
2 · Your own labels (optional)
One label per line. Each is wrapped as “a photo of …” before encoding — that's how CLIP was trained. Include {} to write your own template.
encoding…