Skip to content

BioMysteryBench Human Difficult

Computational biology challenges with objective answers that remained unsolved by independent human experts.

6 models

Higher is better.

  1. 1Gemini 3.8 FlashGoogle56.5
  2. 2Claude Opus 5.5Anthropic50.0
  3. 3Claude Opus 5Anthropic49.4
  4. 4Claude Sonnet 5.5Anthropic44.7
  5. 5Gemini 3.7 FlashGoogle43.5
  6. 6Apodex 1.1Apodex35.3
Higher is better. Bar colour: blue proprietary, orange open weights, green other.

About this benchmark

Metric
Task score
Tasks
Human-difficult 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.