The benchmark for AI skin analysis
SkinBench is an open benchmark for AI skin analysis: compare model accuracy on skin metrics across skin tones, ages and lighting conditions.
Explore the benchmark
Why SkinBench
Diverse dataset
Images across all Fitzpatrick skin types, ages and genders.
Expert annotations
Ground truth labelled by board-certified dermatologists.
Transparent metrics
Open methodology and reproducible scoring for every model.
FAQ
What is SkinBench?
SkinBench is a benchmark that measures how accurately AI models analyse skin from photos.
Who can submit a model?
Any research team or company working on skin analysis can submit a model for evaluation.
Is the benchmark free?
Yes, the benchmark results and methodology are publicly available.