Model

Doubao Seed 2.0 Pro

ByteDance · closed weights

volcengine.com

Coding

Coding
BenchmarkScoreMetricHarnessSourceConfidenceLink
MathVis18.2pass@1N/AIndependentmediumarxiv.org
PremiseSQL0.3exN/AIndependentmediumarxiv.org

Multimodal

Multimodal
BenchmarkScoreMetricHarnessSourceConfidenceLink
FloodVL-Pix0.4clairN/AIndependentmediumarxiv.org
PhysVista91.3order_accuracyN/AIndependentmediumarxiv.org
VisualWorldBench73.8accuracyN/AIndependentmediumarxiv.org

Reproduced by the community

Loading community runs…