Multi-modal Retrieval-Augmented Generation on MC-SEARCH Parallel Image-Text Fork 1.0 (test)
34.83F1 ScoreGemini-2.5-Pro
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| Gemini-2.5-ProAccess=CLOSE, Enhanced=None2026.03 | 34.83 | 34.19 | 2.99 | 16.74 | 1.29 | 53.46 | |
| Qwen2.5-VL-7B + SEARCH-ALIGNAccess=OPEN, Enhanced=SEARCH-ALIGN2026.03 | 32.73 | 24.11 | 1.78 | 25.07 | 1.05 | 55.76 | |
| InternVL3.5-8B + SEARCH-ALIGNAccess=OPEN, Enhanced=SEARCH-ALIGN2026.03 | 30.72 | 24.03 | 2.29 | 23.55 | 1.45 | 58.82 | |
| Gemini-2.5-FlashAccess=CLOSE, Enhanced=None2026.03 | 29.92 | 27.94 | 2.58 | 11.43 | 2.7 | 57.99 | |
| InternVL3.5-8BAccess=OPEN, Enhanced=None2026.03 | 29.74 | 23.05 | 2.13 | 14.22 | 1.75 | 58.82 | |
| Claude-3.7-SonnetAccess=CLOSE, Enhanced=None2026.03 | 26.43 | 24.65 | 1.95 | 14.32 | 1.37 | 58.51 | |
| Qwen2.5-VL-7BAccess=OPEN, Enhanced=None2026.03 | 22.94 | 14.32 | 1.33 | 10.09 | 3.93 | 55.76 | |
| GPT-4o-MiniAccess=CLOSE, Enhanced=None2026.03 | 21.98 | 21.65 | 2.2 | 15 | 1.71 | 53.98 |