ResearchBenchmarksPlan Generation on Commonly solved tasksFollow0.36Runtime RatioDeepSeek0.33880.48190.6250.7681Aug 19, 2025Evaluation ResultsMethodMethodLinksRuntime RatioRuntime F5-3 Success RatePlan Length RatioPlan Length F5-3 Success RateDeepSeekN=1442, Configuration=...N=1442, Configuration=F5-3 (best run)2025.080.36971.161GPTN=1112, Configuration=...N=1112, Configuration=F5-3 (best run)2025.080.42961.5742QwenN=1324, Configuration=...N=1324, Configuration=F5-3 (best run)2025.080.43961.2956LlamaN=1043, Configuration=...N=1043, Configuration=F5-3 (best run)2025.080.89881.3449