← All benchmarks

DrugDiscoveryBench

Science · 2026-06-30

DrugDiscoveryBench evaluates coding agents on 82 expert-curated, verifiable multi-step tasks spanning target identification and validation, hit identification, hit-to-lead analysis, and lead optimization in an adapted Biomni tool environment. The primary metric averages the pass rates of three independent runs per model setting; a task passes only when its final-answer outcome score is 100. The 82-task count is the distinct task cohort, not a single-run binomial denominator for these averaged scores.

Top models (higher is better)