Visual Question Answering on SimpleVQA (Score)
79.6ScoreSimpleSearch-VL-30B-A3B
Evaluation Results
| Method | Links | |
|---|---|---|
| SimpleSearch-VL-30B-A3BEvaluation Paradigm=Multimodal Search Agents2026.06 | 79.6 | |
| GPT-5.2Workflow=Agentic Model, Zero-shot=true2026.07 | 78.18 | |
| SimpleSearch-VL-8BEvaluation Paradigm=Multimodal Search Agents2026.06 | 76.6 | |
| Gemini-3-ProWorkflow=Agentic Model, Zero-shot=true2026.07 | 75.91 | |
| OpenSearch-VL-30B-A3BEvaluation Paradigm=Multimodal Search Agents2026.06 | 74.9 | |
| OpenSearch-VL-8BEvaluation Paradigm=Multimodal Search Agents2026.06 | 71.6 | |
| SenseNova-MARS-8BEvaluation Paradigm=Multimodal Search Agents2026.06 | 70.2 | |
| POINTS-Seeker-8BEvaluation Paradigm=Multimodal Search Agents2026.06 | 68.8 | |
| Gemini-3-FlashWorkflow=Agentic Model, Zero-shot=true2026.07 | 67.92 | |
| VideoSearcher-8BWorkflow=Agentic Model, Zero-shot=false2026.07 | 67.2 | |
| Qwen3-VL-30B-A3BEvaluation Paradigm=Agentic Workflow2026.06 | 66.6 | |
| MM-DeepResearch-8BEvaluation Paradigm=Multimodal Search Agents2026.06 | 65.9 | |
| VideoSearcher-4BWorkflow=Agentic Model, Zero-shot=false2026.07 | 65.84 | |
| Qwen2.5-VL-32B-InstructWorkflow=Agentic Model, Zero-shot=true2026.07 | 65.15 | |
| Qwen3-VL-32B-InstructWorkflow=Agentic Model, Zero-shot=true2026.07 | 64.17 | |
| Gemini-3-ProWorkflow=Direct Answer, Zero-shot=false2026.07 | 64.07 | |
| GPT-4oWorkflow=Agentic Model, Zero-shot=true2026.07 | 63.67 | |
| Gemini-3-FlashWorkflow=Direct Answer, Zero-shot=false2026.07 | 63.57 | |
| Qwen3-VL-8BEvaluation Paradigm=Agentic Workflow2026.06 | 63.4 | |
| Qwen3-VL-8B-InstructWorkflow=Agentic Model, Zero-shot=true2026.07 | 62.29 | |
| GPT-5.2Workflow=Direct Answer, Zero-shot=false2026.07 | 59.92 | |
| DeepEyesV2-7BEvaluation Paradigm=Multimodal Search Agents2026.06 | 59.4 | |
| DeepEyesV2Workflow=Agentic Model, Zero-shot=false2026.07 | 59.4 | |
| WebWatcher-32BEvaluation Paradigm=Multimodal Search Agents2026.06 | 59 | |
| MMSearch-R1-7BEvaluation Paradigm=Multimodal Search Agents2026.06 | 57.4 | |
| MMSearch-R1Workflow=Agentic Model, Zero-shot=false2026.07 | 57.4 | |
| Webwatcher-7BWorkflow=Agentic Model, Zero-shot=false2026.07 | 54.3 | |
| Qwen3-VL-30B-A3BEvaluation Paradigm=Direct Answer2026.06 | 53.2 | |
| GPT-4oWorkflow=Direct Answer, Zero-shot=false2026.07 | 51.73 | |
| Qwen2.5-VL-32B-InstructWorkflow=Direct Answer, Zero-shot=false2026.07 | 48.57 | |
| Qwen2.5-VL-7B-InstructWorkflow=Direct Answer, Zero-shot=false2026.07 | 47.88 | |
| Qwen3-VL-4B-InstructWorkflow=Direct Answer, Zero-shot=false2026.07 | 46.89 | |
| Qwen3-VL-32B-InstructWorkflow=Direct Answer, Zero-shot=false2026.07 | 45.9 | |
| Qwen3-VL-8B-InstructWorkflow=Direct Answer, Zero-shot=false2026.07 | 42.94 | |
| Qwen3-VL-8BEvaluation Paradigm=Direct Answer2026.06 | 42.9 | |
| Visual-ARFTWorkflow=Agentic Model, Zero-shot=false2026.07 | 42.45 | |
| Qwen2.5-VL-7B-InstructWorkflow=Agentic Model, Zero-shot=true2026.07 | 42.35 | |
| Qwen3-VL-4B-InstructWorkflow=Agentic Model, Zero-shot=true2026.07 | 41.95 |