Multimodal Search on MMSearch-Plus (test)
34.5AccuracyGPT-5
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| GPT-5Workflow Protocol=Agentic Workflow2026.06 | 34.5 | — | — | — | |
| Gemini-2.5-ProWorkflow Protocol=Agentic Workflow2026.06 | 30.9 | — | — | — | |
| Visual-SeekerWorkflow Protocol=Multimodal Deep Search Agent2026.06 | 27.3 | — | — | — | |
| Points-Seeker-8BWorkflow Protocol=Multimodal Deep Search Agent2026.06 | 25.2 | — | — | — | |
| Gemini-2.5-FlashWorkflow Protocol=Agentic Workflow2026.06 | 23.3 | — | — | — | |
| Claude-4-SonnetWorkflow Protocol=Agentic Workflow2026.06 | 20.9 | — | — | — | |
| Vision-DeepResearch-8BWorkflow Protocol=Multimodal Deep Search Agent2026.06 | 20.4 | — | — | — | |
| GPT-5Workflow Protocol=Direct Answer2026.06 | 19.1 | — | — | — | |
| Gemini-2.5-ProWorkflow Protocol=Direct Answer2026.06 | 14.5 | — | — | — | |
| Qwen3-VL-8B-InstructWorkflow Protocol=Agentic Workflow2026.06 | 10.9 | — | — | — | |
| Gemini-2.5-FlashWorkflow Protocol=Direct Answer2026.06 | 8.1 | — | — | — | |
| Claude-4-SonnetWorkflow Protocol=Direct Answer2026.06 | 4 | — | — | — | |
| Qwen3-VL-8B-InstructWorkflow Protocol=Direct Answer2026.06 | 3.2 | — | — | — | |
| Agent Workflow MemoryModel=Qwen3-VL-235B, Setting=Agent Workflow Memory2026.03 | — | 12.32 | 20.38 | 4.03 | |
| Agent Workflow MemoryModel=Qwen3-VL-32B, Setting=Agent Workflow Memory2026.03 | — | 9.48 | 18.48 | 2.77 | |
| Agent-KBModel=Qwen3-VL-235B, Setting=Agent-KB2026.03 | — | 10.29 | 17.54 | 4.71 | |
| Agent-KBModel=Qwen3-VL-32B, Setting=Agent-KB2026.03 | — | 11.37 | 22.27 | 3.04 | |
| Dynamic CheatSheetModel=Qwen3-VL-235B, Setting=Dynamic CheatSheet2026.03 | — | 10.82 | 18.01 | 3.87 | |
| Dynamic CheatSheetModel=Qwen3-VL-32B, Setting=Dynamic CheatSheet2026.03 | — | 11.14 | 21.8 | 3.13 | |
| No ToolsModel=Qwen3-VL-235B, Setting=No Tools2026.03 | — | 2.49 | 6.16 | — | |
| No ToolsModel=Qwen3-VL-32B, Setting=No Tools2026.03 | — | 2.37 | 4.74 | — | |
| w/ ToolsModel=Qwen3-VL-235B, Setting=w/ Tools2026.03 | — | 8.76 | 14.69 | 3.83 | |
| w/ ToolsModel=Qwen3-VL-32B, Setting=w/ Tools2026.03 | — | 9.95 | 19.91 | 2.88 | |
| XSkillModel=Qwen3-VL-235B, Setting=XSkill2026.03 | — | 10.43 | 21.8 | 4.15 | |
| XSkillModel=Qwen3-VL-32B, Setting=XSkill2026.03 | — | 13.1 | 24.17 | 4.18 |