QvQ-72B-Preview

An experimental research model focusing on advanced visual reasoning and step-by-step cognitive capabilities. Achieves strong performance on multi-modal science and mathematics tasks, though exhibits some limitations such as potential language mixing and recursive reasoning loops.

Benchmark results

Benchmark Score Tags Source
MathVision 35.9% self-reported llm-stats link →
MathVista 71.4% self-reported llm-stats link →
MMMU 70.3% self-reported llm-stats link →
OlympiadBench 20.4% self-reported llm-stats link →