Where the numbers come from
Data transparency
Every price and every ranking position on this site is a function of counted credentials. This page lists the APIs and registries that feed those counts, the exact weight each credential carries, and the source passages the AI relied on when building an artist CV. Nothing here is an appraisal.
01 · Coverage of the current database
Artists
…
With declared sources
…
CVs with evidence
…
Average confidence
…
No AI refresh recorded yet. Records refresh automatically every day and can be refreshed on demand from the ranking.
02 · What feeds a record
The research pass consults these registries, then reconciles them into the counted credentials stored for each artist.
Metropolitan Museum of Art — Collection API
Open object endpoint used to confirm works held in the collection.
Feeds: Museum collections
Harvard Art Museums API
Person and exhibition records for institutional history.
Feeds: Museum collections · Exhibitions
Wikidata / Wikipedia
Identity data, birth and death years, practice and mediums.
Feeds: Biography · Nationality · Dates · Mediums
Gallery and institution websites
Representation pages and exhibition archives published by the venues.
Feeds: Gallery representation · Solo and group exhibitions
Auction house results
Published hammer results — reference only, never part of the cofactor.
Feeds: Auction record
Lovable AI research pass
The model reconciles the sources above into counted credentials and a structured CV, and reports its own confidence.
Feeds: Counting, reconciliation and CV
Manual review screen
Values corrected by hand on the artist credential editor override the research pass.
Feeds: Any credential
03 · Sources actually cited
Registries the AI declared for the artists currently in the database, and which credentials each one supported.
Loading records…
04 · How a credential becomes a price
Each credential count is multiplied by its weight; the sum is the credential score. Cofactor = 1.5 + 2.4 × √score. Painting price = height (in) × width (in) × cofactor. Photography uses market precedent by size bracket and edition scarcity. The auction record is shown for reference and never enters the cofactor.
05 · AI evidence in each CV
When a CV is generated, the AI also records the passages it relied on — the section, the entry they justify, the quoted material and the source. Open any artist from the ranking and read the “Transparency” block at the bottom of the panel. Entries generated before this feature was added show no passages until the CV is refreshed.
Estimated values in USD, derived from public data gathered by AI. Use as a reference, not as a formal appraisal.