ResearchBenchmarksPrompt Optimization on GEPA Evaluation Suite AggregateFollow62.02Aggregate ScoreLEVI48.323251.879155.43558.9909May 10, 2026Evaluation ResultsMethodMethodLinksAggregate ScoreImprovementOptimization Budget (Rollouts)LEVITask Model=Qwen 3 8BTask Model=Qwen 3 8B2026.0562.0213.172,191GEPATask Model=Qwen 3 8BTask Model=Qwen 3 8B2026.0561.2812.444,985MIPROv2Task Model=Qwen 3 8BTask Model=Qwen 3 8B2026.0555.116.26—BaselineTask Model=Qwen 3 8BTask Model=Qwen 3 8B2026.0548.85——