ZeroBench

Study / Research

A challenging benchmark that tests AI models on tasks like reading and summing prices from a misread menu in poor lighting.

Mentioned in 1 video

Videos Mentioning ZeroBench

Opus 5.5: How Close Are We to Automated AI Research?

Opus 5.5: How Close Are We to Automated AI Research?

AI Explained

A challenging benchmark that tests AI models on tasks like reading and summing prices from a misread menu in poor lighting.