Unknown
mmbench is a state-of-the-art machine learning benchmark indexed on Codesota. This page tracks published model results, top scores per metric, and the SOTA timeline for mmbench.
Higher is better
| Rank | Model | Source | Score | Year | Paper |
|---|---|---|---|---|---|
| 1 | Qwen2.5-VL 72B MMBench EN test. Qwen2.5-VL 72B. Table 2. arxiv:2502.13923 | Community | 90.5 | 2026 | Source |
| 2 | InternVL3-78B MMBench EN test. InternVL3-78B. Table 2. arxiv:2501.12891 | Community | 90.1 | 2026 | Source |
| 3 | Qwen2-VL 72B MMBench EN test. Qwen2-VL 72B. Table 6. arxiv:2409.12191 | Community | 88 | 2026 | Source |
| 4 | InternVL2-76B MMBench EN test. InternVL2-76B. Table 12. arxiv:2404.16821 | Community | 86.5 | 2026 | Source |
| 5 | GPT-4o MMBench EN test. GPT-4o. System card Table 1. arxiv:2410.21276 | Community | 83.4 | 2026 | Source |
| 6 | GPT-4V MMBench EN test. GPT-4V. Reported in multiple comparison papers incl. InternVL2 Table 12. | Community | 75.8 | 2026 | Source |
| 7 | Gemini 1.5 Pro MMBench EN dev. Gemini 1.5 Pro. Table 5. arxiv:2403.05530 | Community | 73.9 | 2026 | Source |
| 8 | LLaVA-1.5 MMBench EN dev. 13B. Table 1. arxiv:2310.03744 | Community | 67.7 | 2026 | Source |