Benchmark

LABBench2

LABBench2 evaluates models on real-world biology research tasks.

Modality
text
Categories
reasoning, science, agents, biology
Openness
unknown
Source
llm_stats
Reported scores
1

Reported scores

llm_stats

ModelOrganizationReported valueReportedEvidence
Gemini 3.7 FlashGoogle0.8212026-08-13evidence

Scores are partitioned by the source that reported them and are never merged into a single cross-source ranking, because the sources measure different things and say so.