Skip to content

BioMysteryBench Human Solvable

Computational biology challenges that independent human experts were able to solve.

5 models

Higher is better.

  1. 1Claude Opus 5Anthropic90.1
  2. 2Claude Opus 5.5Anthropic89.3
  3. 3Claude Sonnet 5.5Anthropic89.2
  4. 4Gemini 3.8 FlashGoogle88.8
  5. 5Gemini 3.7 FlashGoogle87.1
Higher is better. Bar colour: blue proprietary, orange open weights, green other.

About this benchmark

Metric
Task score
Tasks
Human-solvable computational biology investigations
Authors
Anthropic (2026)

Scores are as reported by model developers or evaluators and may use different settings (for example reasoning effort or tools). Compare with care.