Deep Search on GAIA text-only (val)
70.9AccuracyTongyi-DeepResearch
Evaluation Results
| Method | Links | |
|---|---|---|
| Tongyi-DeepResearchContext Budget=128k2026.01 | 70.9 | |
| OpenAI-o3Context Budget=128k2026.01 | 70.5 | |
| GLM-4.5Context Budget=128k2026.01 | 66 | |
| DeepSeek-v3.1Context Budget=128k2026.01 | 63.1 | |
| DeepDive-30B-SFT + C-GRPOContext Budget=128k2026.01 | 56.3 | |
| DeepDive-30B-SFT + E-GRPOContext Budget=128k2026.01 | 55.3 | |
| DeepDive-30B-SFT + C-GRPOContext Budget=64k2026.01 | 53.7 | |
| Asearcher-Web-32BContext Budget=128k2026.01 | 52.8 | |
| DeepDive-30B-SFT + E-GRPOContext Budget=64k2026.01 | 52.8 | |
| WebSailor-32BContext Budget=32k2026.01 | 51.5 | |
| DeepDive-30B-SFT + GRPOContext Budget=64k2026.01 | 51.1 | |
| DeepDive-30B-SFT + GRPOContext Budget=128k2026.01 | 51.1 | |
| DeepDive-30B-SFTContext Budget=128k2026.01 | 50.8 | |
| DeepDive-4B-SFT + C-GRPOContext Budget=128k2026.01 | 50.2 | |
| WebExplorer-8BContext Budget=128k2026.01 | 50 | |
| DeepDive-4B-SFT + C-GRPOContext Budget=64k2026.01 | 48.9 | |
| DeepDive-4B-SFTContext Budget=128k2026.01 | 46 | |
| DeepDive-30B-SFTContext Budget=64k2026.01 | 46 | |
| DeepDive-4B-SFT + E-GRPOContext Budget=64k2026.01 | 42.4 | |
| DeepDive-4B-SFT + E-GRPOContext Budget=128k2026.01 | 42.4 | |
| DeepDive-4B-SFT + GRPOContext Budget=128k2026.01 | 41.1 | |
| DeepDive-4B-SFT + GRPOContext Budget=64k2026.01 | 40.5 | |
| DeepDive-4B-SFTContext Budget=64k2026.01 | 39.5 | |
| WebSailor-7BContext Budget=32k2026.01 | 37.9 |