Agentic Search on GAIA Search 2
82.71Pass@1OpenAI GPT-5.4
Evaluation Results
| Method | Links | |
|---|---|---|
| OpenAI GPT-5.4runs=32026.06 | 82.71 | |
| Gemini 3.1-Proruns=32026.06 | 80.83 | |
| DS-V4-Proruns=32026.06 | 78.33 | |
| Ring-2.6-1Tvariant=xhigh, runs=32026.06 | 77.9 | |
| Kimi-K2.6runs=32026.06 | 76.04 | |
| GLM-5.1runs=32026.06 | 75.63 | |
| Ring-2.6-1Tvariant=high, runs=32026.06 | 75.4 | |
| Claude Opus-4.6runs=32026.06 | 73.75 | |
| Claude Opus-4.7runs=32026.06 | 67.92 |