Model

Qwen2.5-VL-3B

Alibaba · open weights

qwenlm.github.io

Multimodal

Multimodal
BenchmarkScoreMetricHarnessSourceConfidenceLink
OmniVCBench22.5accuracyN/AIndependentmediumarxiv.org
P^3IBench16.7overall_average_prs-bN/AIndependentmediumarxiv.org
ViSiP-Bench63.2parN/AIndependentmediumarxiv.org

Reproduced by the community

Loading community runs…