About

A unified, measured index of the world's open-access art collections.

Where the data comes from

Metadata and images come directly from the open-access programmes of National Gallery of Art, The Cleveland Museum of Art, The Metropolitan Museum of Art, Smithsonian Institution and National Gallery of Denmark (SMK). Nothing is scraped: each museum publishes an API or bulk dataset for exactly this purpose.

What is added

Two things, both computed from the image files themselves. Colour palettes are extracted by clustering pixels in CIE L*a*b* space, which matches human perception far better than RGB, then matched to the nearest of 48 named tones. Maximum print size is the pixel dimensions divided by 300 DPI, the density below which a print starts to look soft.

What is not here

No generated descriptions, no invented history, no AI-written commentary. If a fact appears on this site it came from a museum record or from measuring the file. Where we have nothing to add, we say nothing.