Model

MiMo V2.6 Pro

Xiaomi · open weights

github.com

Knowledge & Reasoning

Knowledge & Reasoning
BenchmarkScoreMetricHarnessSourceConfidenceLink
CritPt26.6accuracyN/AIndependentlowepoch.ai

Coding

Coding
BenchmarkScoreMetricHarnessSourceConfidenceLink
SciCode60.9scoreN/AIndependentmediumartificialanalysis.ai

Agentic / Tool Use

Agentic / Tool Use
BenchmarkScoreMetricHarnessSourceConfidenceLink
BioStudyBench66.7pass_rateN/AIndependentmediumarxiv.org

Multimodal

Multimodal
BenchmarkScoreMetricHarnessSourceConfidenceLink
EMBER-Bench26.3accuracyN/AIndependentmediumarxiv.org
GeoSpatialBench63.0accuracyN/AIndependentmediumarxiv.org

Reproduced by the community

Loading community runs…