Model

Qwen3.5 122B A10B

Alibaba · open weights

qwenlm.github.io

Knowledge & Reasoning

Knowledge & Reasoning
BenchmarkScoreMetricHarnessSourceConfidenceLink
CritPt0.9accuracyN/AIndependentlowepoch.ai

Coding

Coding
BenchmarkScoreMetricHarnessSourceConfidenceLink
SciCode35.6scoreN/AIndependentmediumartificialanalysis.ai

Multimodal

Multimodal
BenchmarkScoreMetricHarnessSourceConfidenceLink
EgoGears36.8accuracyN/AIndependentmediumarxiv.org
EMBER-Bench37.1accuracyN/AIndependentmediumarxiv.org
VisualNoiseQA54.6accuracybpIndependentmediumarxiv.org

Multilingual

Multilingual
BenchmarkScoreMetricHarnessSourceConfidenceLink
TRuST42.7recallt-searchIndependentmediumarxiv.org

Reproduced by the community

Loading community runs…