Agentic Search on GAIA Search
96.9AccuracyWeighted Voting
Evaluation Results
| Method | Links | |
|---|---|---|
| Weighted VotingModel=Gemini-3-flash, Sampling Trajectories=42026.05 | 96.9 | |
| Best-of-NModel=Gemini-3-flash, Sampling Trajectories=42026.05 | 96.9 | |
| Solution AggregationModel=Gemini-3-flash, Sampling Trajectories=42026.05 | 95.3 | |
| Confidence VerifyModel=Gemini-3-flash, Sampling Trajectories=42026.05 | 95.3 | |
| FINEVERIFYModel=Gemini-3-flash, Sampling Trajectories=42026.05 | 95.3 | |
| Majority VotingModel=Gemini-3-flash, Sampling Trajectories=42026.05 | 93.8 | |
| Pass@1Model=Gemini-3-flash, Sampling Trajectories=12026.05 | 89 | |
| FINEVERIFYModel=GPT-5-mini, Sampling Trajectories=42026.05 | 81.3 | |
| Solution AggregationModel=GPT-5-mini, Sampling Trajectories=42026.05 | 79.7 | |
| Confidence VerifyModel=GPT-5-mini, Sampling Trajectories=42026.05 | 79.7 | |
| Majority VotingModel=GPT-5-mini, Sampling Trajectories=42026.05 | 78.1 | |
| Pass@1Model=GPT-5-mini, Sampling Trajectories=12026.05 | 76.6 | |
| Weighted VotingModel=GPT-5-mini, Sampling Trajectories=42026.05 | 76.6 | |
| Best-of-NModel=GPT-5-mini, Sampling Trajectories=42026.05 | 75 |