Information Retrieval on Web synthetic
81.9RecallOpus-4.6
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Opus-4.6Category=Frontier models2026.06 | 81.9 | 92.8 | 83.2 | |
| GPT-5.4Category=Frontier models2026.06 | 80 | 85.8 | 82.1 | |
| Harness-1Parameters=20B, Category=Open-source small models2026.06 | 78.7 | 84.5 | 88.1 | |
| Sonnet-4.6Category=Frontier models2026.06 | 74.4 | 89.2 | 76.7 | |
| Kimi-K2.5Category=Frontier models2026.06 | 70.7 | 84.3 | 87.8 | |
| Context-1Parameters=20B, Category=Open-source small models2026.06 | 68.9 | 77.5 | 82.3 | |
| Tongyi DeepResearchParameters=30B, Category=Open-source small models2026.06 | 60.7 | 67.9 | 72.2 | |
| GPT-OSSParameters=120B, Category=Frontier models2026.06 | 47.8 | 66.4 | 79.6 | |
| GPT-OSSParameters=20B, Category=Open-source small models2026.06 | 17.4 | 25.2 | 51.2 | |
| Search-R1Parameters=32B, Category=Open-source small models2026.06 | 13.5 | 24 | 13.5 | |
| Qwen3Parameters=32B, Category=Open-source small models2026.06 | 13.2 | 21.1 | 41 |