Web-search QA on BrowseComp-VL
54.9Pass@1GPT-5
Evaluation Results
| Method | Links | |
|---|---|---|
| GPT-5Workflow=RAG Workflow2026.05 | 54.9 | |
| GPT-5Workflow=Direct Reasoning2026.05 | 48.6 | |
| OpenSearch-VL-32BWorkflow=Agentic Workflow2026.05 | 43.8 | |
| Gemini-2.5-ProWorkflow=Direct Reasoning2026.05 | 43.1 | |
| OpenSearch-VL-30B-A3BWorkflow=Agentic Workflow2026.05 | 41.1 | |
| OpenSearch-VL-8BWorkflow=Agentic Workflow2026.05 | 37.6 | |
| Gemini-2.5-FlashWorkflow=Direct Reasoning2026.05 | 37.1 | |
| SenseNova-MARS-8BWorkflow=Agentic Workflow2026.05 | 35.1 | |
| Qwen3-VL-32BWorkflow=Agentic Workflow2026.05 | 35.1 | |
| Qwen3-VL-30B-A3BWorkflow=Agentic Workflow2026.05 | 34.1 | |
| Claude-3.7-SonnetWorkflow=Direct Reasoning2026.05 | 32.3 | |
| Qwen3-VL-32BWorkflow=Direct Reasoning2026.05 | 30.8 | |
| Qwen3-VL-30B-A3BWorkflow=Direct Reasoning2026.05 | 29.6 | |
| Claude-4-SonnetWorkflow=Direct Reasoning2026.05 | 29.3 | |
| Qwen3-VL-8BWorkflow=RAG Workflow2026.05 | 29.3 | |
| Qwen3-VL-8BWorkflow=Agentic Workflow2026.05 | 27.9 | |
| WebWatcher-32BWorkflow=Agentic Workflow2026.05 | 27 | |
| Qwen3-VL-8BWorkflow=Direct Reasoning2026.05 | 24.1 | |
| WebWatcher-7BWorkflow=Agentic Workflow2026.05 | 21.2 | |
| MMSearch-R1-7BWorkflow=Agentic Workflow2026.05 | 20.9 | |
| Visual-ARFT-7BWorkflow=Agentic Workflow2026.05 | 16.5 | |
| GPT-4oWorkflow=RAG Workflow2026.05 | 13.4 | |
| Claude-3.7-SonnetWorkflow=RAG Workflow2026.05 | 10 | |
| GPT-4oWorkflow=Direct Reasoning2026.05 | 5.5 |