ResearchBenchmarksMathematical Reasoning on AIME 24 (REST, RSTAR, REBASE, ETS, AVG)Follow76.7REST ScoreSPEX69.73271.54173.3575.159May 11, 2026Evaluation ResultsMethodMethodLinksREST ScoreRSTAR ScoreREBASE ScoreETS ScoreAverage Score (AVG)SPEXModel=DeepseekModel=Deepseek2026.0576.766.776.776.773.8BaselineModel=DeepseekModel=Deepseek2026.0573.366.78073.373.3SPEXModel=QwenModel=Qwen2026.0573.38076.776.777.9BaselineModel=QwenModel=Qwen2026.0570708073.373.8