Multimodal Search on LiveVQA
84.1AccuracyHyperEyes-235B
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| HyperEyes-235BTraining Stage=RL, Evaluation Mode=Agentic Workflow2026.05 | 84.1 | 1.5 | |
| HyperEyes-235BTraining Stage=SFT, Evaluation Mode=Agentic Workflow2026.05 | 83.7 | 2.1 | |
| HyperEyes-30BTraining Stage=RL, Evaluation Mode=Agentic Workflow2026.05 | 81.6 | 1.7 | |
| HyperEyes-30BTraining Stage=SFT, Evaluation Mode=Agentic Workflow2026.05 | 80.3 | 1.9 | |
| REDSearchEvaluation Mode=Multimodal Deep Search Agents2026.05 | 79.3 | — | |
| VDREvaluation Mode=Multimodal Deep Search Agents2026.05 | 77.6 | 10.2 | |
| Kimi-K2.5Evaluation Mode=Agentic Workflow2026.05 | 76.6 | 2.1 | |
| Gemini-3.1-ProEvaluation Mode=Agentic Workflow2026.05 | 76.6 | 1.4 | |
| Claude-Opus-4.6Evaluation Mode=Agentic Workflow2026.05 | 67.4 | 1.2 | |
| WebWatcherEvaluation Mode=Multimodal Deep Search Agents2026.05 | 58.7 | 4.1 | |
| Qwen3-VL-235BEvaluation Mode=Agentic Workflow2026.05 | 58.2 | 1.6 | |
| DeepEyes-V2Evaluation Mode=Multimodal Deep Search Agents2026.05 | 58 | 3.7 | |
| Kimi-K2.5Evaluation Mode=Direct Answer2026.05 | 57.3 | — | |
| Claude-Opus-4.6Evaluation Mode=Direct Answer2026.05 | 53.1 | — | |
| Gemini-3.1-ProEvaluation Mode=Direct Answer2026.05 | 51.5 | — | |
| Qwen3-VL-30BEvaluation Mode=Agentic Workflow2026.05 | 49.8 | 1.9 | |
| MMSearch-R1Evaluation Mode=Multimodal Deep Search Agents2026.05 | 48.4 | 1.4 | |
| Qwen3-VL-235BEvaluation Mode=Direct Answer2026.05 | 41.4 | — | |
| Qwen3-VL-30BEvaluation Mode=Direct Answer2026.05 | 35.6 | — |