BioMysteryBench Human-Difficult

Anthropic BioMysteryBench slice covering 23 real-world bioinformatics tasks no human benchmarker solved after QC, evaluated by average accuracy across five trials per problem.

9rows
accuracyprimary metric
2026-05-28sampled

Metadata

ID
biomysterybench_human_difficult
Category
Science
Release
2026-04-29

Metrics

Accuracy

Showing 2 latest source slices.

Latest Results

Provider-published system-card benchmark scores parsed from Anthropic's Claude Opus 4.8 capability evaluation tables. Rows are marked self-reported and should be interpreted as source claims unless independently reproduced.

Rank Subject Accuracy Model Match Provenance Sampled
1 Claude Opus 4.8 40% Claude Opus 4.8
anthropic-claude-opus-4.8
 		Self-reported 2026-05-28
2 Claude Mythos Preview 29.6% Claude Mythos Preview
anthropic-claude-mythos-preview
 		Self-reported 2026-05-28
3 Claude Opus 4.7 24.7% Claude Opus 4.7
anthropic-claude-opus-4.7
 		Self-reported 2026-05-28
4 Claude Sonnet 4.6 19.1% Claude Sonnet 4.6
anthropic-claude-sonnet-4.6
 		Self-reported 2026-05-28
1 Claude Mythos Preview 29.6% Claude Mythos Preview
anthropic-claude-mythos-preview
 		Self-reported 2026-04-29
2 Claude Opus 4.7 27.0% Claude Opus 4.7
anthropic-claude-opus-4.7
 		Self-reported 2026-04-29
3 Claude Opus 4.6 23.5% Claude Opus 4.6
anthropic-claude-opus-4.6
 		Self-reported 2026-04-29
4 Claude Sonnet 4.6 19.1% Claude Sonnet 4.6
anthropic-claude-sonnet-4.6
 		Self-reported 2026-04-29
5 Claude Haiku 4.5 5.2% Claude Haiku 4.5
anthropic-claude-haiku-4.5
 		Self-reported 2026-04-29