Multimodal Question Answering on MMBench
83.6MMBench ScoreQwen3-VL
Evaluation results
| Method | Links | |
|---|---|---|
| Qwen3-VLScale=4B2026.05 | 83.6 | |
| Qwen3.5-VLScale=4B2026.05 | 81.8 | |
| Qwen3.5-VLScale=2B2026.05 | 78 | |
| DatologyScale=4B2026.05 | 76.7 | |
| InternVL3.5Scale=2B2026.05 | 75.7 | |
| Qwen3-VLScale=2B2026.05 | 75.7 | |
| MAmmoTH-VLScale=4B2026.05 | 74.4 | |
| InternVL3Scale=2B2026.05 | 72.2 | |
| DatologyScale=2B2026.05 | 70.3 | |
| MAmmoTH-VLScale=2B2026.05 | 69.7 | |
| InternVL3.5Scale=4B2026.05 | 68.6 | |
| DatologyScale=1B2026.05 | 63.3 | |
| MAmmoTH-VLScale=1B2026.05 | 61.5 |