Multimodal Understanding on MMBench Chinese
89.5MMB Benchmark (CN)Qwen3-VL-30B A3B-Instruct
Evaluation results
| Method | Links | |
|---|---|---|
| Qwen3-VL-30B A3B-InstructParameters=30B A3B, Type=Instruct2025.12 | 89.5 | |
| Qwen3-VL 8B-InstructParameters=8B, Type=Instruct2025.12 | 88 | |
| Step-GUI-8BParameters=8B2025.12 | 88 | |
| Step-GUI-4BParameters=4B2025.12 | 87 | |
| GLM-4.6V FlashParameters=4.6B, Variant=Flash2025.12 | 85.8 | |
| InternVL3.5-8BRatio=100%, Backbone=InternVL3.5-8B2026.05 | 83.07 | |
| VanillaBackbone=InternVL-2.5-8B, Token Reduction Ratio=0%2026.06 | 82.6 | |
| CLSE globalBackbone=InternVL-2.5-8B, Token Reduction Ratio=66.7%, FFT Type=Global2026.06 | 82.6 | |
| CLSE localBackbone=InternVL-2.5-8B, Token Reduction Ratio=66.7%, FFT Type=Local2026.06 | 82.5 | |
| F3ARatio=60%, Backbone=InternVL3.5-8B2026.05 | 82.49 | |
| FastVBackbone=InternVL-2.5-8B, Token Reduction Ratio=66.7%2026.06 | 82.4 | |
| VisionZipRatio=60%, Backbone=InternVL3.5-8B2026.05 | 82.24 | |
| GPT5-nano minimalVariant=nano minimal2025.12 | 82.2 | |
| CDPrunerRatio=60%, Backbone=InternVL3.5-8B2026.05 | 82.09 | |
| CLSE localBackbone=InternVL-2.5-8B, Token Reduction Ratio=77.8%, FFT Type=Local2026.06 | 82 | |
| F3ARatio=40%, Backbone=InternVL3.5-8B2026.05 | 81.82 | |
| CLSE globalBackbone=InternVL-2.5-8B, Token Reduction Ratio=77.8%, FFT Type=Global2026.06 | 81.8 | |
| DivPruneRatio=60%, Backbone=InternVL3.5-8B2026.05 | 81.66 | |
| FastVRatio=60%, Backbone=InternVL3.5-8B2026.05 | 81.57 | |
| FastVRatio=40%, Backbone=InternVL3.5-8B2026.05 | 81.08 | |
| CDPrunerRatio=40%, Backbone=InternVL3.5-8B2026.05 | 81.01 | |
| VisionZipRatio=40%, Backbone=InternVL3.5-8B2026.05 | 80.83 | |
| DivPruneRatio=40%, Backbone=InternVL3.5-8B2026.05 | 80.58 | |
| FastVBackbone=InternVL-2.5-8B, Token Reduction Ratio=77.8%2026.06 | 80.5 | |
| CLSE globalBackbone=InternVL-2.5-8B, Token Reduction Ratio=88.9%, FFT Type=Global2026.06 | 80.5 | |
| CLSE localBackbone=InternVL-2.5-8B, Token Reduction Ratio=88.9%, FFT Type=Local2026.06 | 80.3 | |
| F3ARatio=20%, Backbone=InternVL3.5-8B2026.05 | 79.5 | |
| IXC2-VLLLM=InernLM2-7B2024.04 | 79.4 | |
| VT-LMMLLM=Mistral-7B, Res.=3362024.03 | 79 | |
| VisionZipRatio=20%, Backbone=InternVL3.5-8B2026.05 | 78.5 | |
| CDPrunerRatio=20%, Backbone=InternVL3.5-8B2026.05 | 78.06 | |
| IXC2-4KHDLLM=InernLM2-7B2024.04 | 77.7 | |
| DivPruneRatio=20%, Backbone=InternVL3.5-8B2026.05 | 77.67 | |
| FastVRatio=20%, Backbone=InternVL3.5-8B2026.05 | 76.84 | |
| FastVBackbone=InternVL-2.5-8B, Token Reduction Ratio=88.9%2026.06 | 74.4 | |
| LLaVA-XTunerLLM=InernLM2-20B2024.04 | 73.7 | |
| InternLM-XCLLM=InernLM-7B2024.04 | 72.4 | |
| LLaVA-NextLLM=Vicuna-13B2024.04 | 68.5 | |
| MonkeyLLM=Qwen-7B2024.04 | 67.5 | |
| LLaVA-1.5LLM=Vicuna-13B2024.04 | 63.6 | |
| LLaVA-1.5LLM Backbone=Vicuna-13B2023.12 | 63.6 | |
| LLaVA-v1.5LLM=Vicuna-13B, Res.=3362024.03 | 63.6 | |
| LLaVA-1.5-13BBackbone=LLaVA-1.5-13B, Retained Tokens=576, Pruning Ratio=0%2026.02 | 63.5 | |
| FSRBackbone=LLaVA-NeXT-13B, Retained Tokens=960, Pruning Ratio=66.7%2026.02 | 63.3 | |
| FSRBackbone=LLaVA-NeXT-13B, Retained Tokens=640, Pruning Ratio=77.8%2026.02 | 63.3 | |
| VisPrunerBackbone=LLaVA-NeXT-13B, Retained Tokens=960, Pruning Ratio=66.7%2026.02 | 63.1 | |
| FSRBackbone=LLaVA-NeXT-13B, Retained Tokens=320, Pruning Ratio=88.9%2026.02 | 63.1 | |
| FSRBackbone=LLaVA-1.5-13B, Retained Tokens=192, Pruning Ratio=66.7%2026.02 | 63 | |
| DivPruneBackbone=LLaVA-NeXT-13B, Retained Tokens=640, Pruning Ratio=77.8%2026.02 | 62.9 | |
| VisionZipBackbone=LLaVA-NeXT-13B, Retained Tokens=640, Pruning Ratio=77.8%2026.02 | 62.6 | |
| VisPrunerBackbone=LLaVA-NeXT-13B, Retained Tokens=640, Pruning Ratio=77.8%2026.02 | 62.6 | |
| VisPrunerBackbone=LLaVA-1.5-13B, Retained Tokens=192, Pruning Ratio=66.7%2026.02 | 62.5 | |
| VisionZipBackbone=LLaVA-1.5-13B, Retained Tokens=128, Pruning Ratio=77.8%2026.02 | 62.5 | |
| CDPrunerBackbone=LLaVA-NeXT-13B, Retained Tokens=960, Pruning Ratio=66.7%2026.02 | 62.5 | |
| CDPrunerBackbone=LLaVA-NeXT-13B, Retained Tokens=640, Pruning Ratio=77.8%2026.02 | 62.5 | |
| HoloVBackbone=LLaVA-NeXT-13B, Retained Tokens=320, Pruning Ratio=88.9%2026.02 | 62.5 | |
| VisPrunerBackbone=LLaVA-NeXT-13B, Retained Tokens=320, Pruning Ratio=88.9%2026.02 | 62.5 | |
| LyricsLLM Backbone=Vicuna-13B2023.12 | 62.4 | |
| FastVBackbone=LLaVA-1.5-13B, Retained Tokens=128, Pruning Ratio=77.8%2026.02 | 62.3 | |
| VisionZipBackbone=LLaVA-NeXT-13B, Retained Tokens=320, Pruning Ratio=88.9%2026.02 | 62.3 | |
| ShareGPT4VLLM=Vicuna-7B2024.04 | 62.2 | |
| ShareGPT4VLLM Backbone=Vicuna-7B2023.12 | 62.2 | |
| VisPrunerBackbone=LLaVA-1.5-13B, Retained Tokens=128, Pruning Ratio=77.8%2026.02 | 62.2 | |
| CDPrunerBackbone=LLaVA-1.5-13B, Retained Tokens=192, Pruning Ratio=66.7%2026.02 | 62.1 | |
| HoloVBackbone=LLaVA-1.5-13B, Retained Tokens=128, Pruning Ratio=77.8%2026.02 | 62 | |
| DivPruneBackbone=LLaVA-NeXT-13B, Retained Tokens=320, Pruning Ratio=88.9%2026.02 | 61.9 | |
| CDPrunerBackbone=LLaVA-NeXT-13B, Retained Tokens=320, Pruning Ratio=88.9%2026.02 | 61.9 | |
| FSRBackbone=LLaVA-1.5-13B, Retained Tokens=128, Pruning Ratio=77.8%2026.02 | 61.6 | |
| CDPrunerBackbone=LLaVA-1.5-13B, Retained Tokens=128, Pruning Ratio=77.8%2026.02 | 61.4 | |
| VisionZipBackbone=LLaVA-1.5-13B, Retained Tokens=64, Pruning Ratio=88.9%2026.02 | 61.3 | |
| LLaVA-NeXT-13BBackbone=LLaVA-NeXT-13B, Retained Tokens=2880, Pruning Ratio=0%2026.02 | 61.2 | |
| DivPruneBackbone=LLaVA-1.5-13B, Retained Tokens=128, Pruning Ratio=77.8%2026.02 | 60.7 | |
| FSRBackbone=LLaVA-1.5-13B, Retained Tokens=64, Pruning Ratio=88.9%2026.02 | 60.7 | |
| HoloVBackbone=LLaVA-NeXT-13B, Retained Tokens=640, Pruning Ratio=77.8%2026.02 | 60.6 | |
| HoloVBackbone=LLaVA-1.5-13B, Retained Tokens=64, Pruning Ratio=88.9%2026.02 | 60.3 | |
| HoloVBackbone=LLaVA-1.5-13B, Retained Tokens=192, Pruning Ratio=66.7%2026.02 | 60.2 | |
| VisPrunerBackbone=LLaVA-1.5-13B, Retained Tokens=64, Pruning Ratio=88.9%2026.02 | 59.9 | |
| FastVBackbone=LLaVA-NeXT-13B, Retained Tokens=640, Pruning Ratio=77.8%2026.02 | 59.9 | |
| VT-LMMLLM=Vicuna-7B, Res.=3362024.03 | 59.8 | |
| DivPruneBackbone=LLaVA-1.5-13B, Retained Tokens=64, Pruning Ratio=88.9%2026.02 | 59.8 | |
| CDPrunerBackbone=LLaVA-1.5-13B, Retained Tokens=64, Pruning Ratio=88.9%2026.02 | 58.9 | |
| LLaVA-v1.5LLM=Vicuna-7B, Res.=3362024.03 | 58.3 | |
| Qwen-VL-ChatLLM=Qwen-7B2024.04 | 56.7 | |
| Qwen-VL-ChatLLM Backbone=Qwen-7B2023.12 | 56.7 | |
| Qwen-VL-ChatLLM=Qwen-7B, Res.=4482024.03 | 56.7 | |
| HoloVBackbone=LLaVA-NeXT-13B, Retained Tokens=960, Pruning Ratio=66.7%2026.02 | 56.4 | |
| CogVLM-17BLLM=Vicuna-7B2024.04 | 55.9 | |
| FastVBackbone=LLaVA-1.5-13B, Retained Tokens=64, Pruning Ratio=88.9%2026.02 | 55.1 | |
| FastVBackbone=LLaVA-NeXT-13B, Retained Tokens=320, Pruning Ratio=88.9%2026.02 | 54.4 | |
| IDEFICS-80BLLM Backbone=LLaMA-65B2023.12 | 38.1 | |
| IDEFICS-80BLLM=LLAMA-65B, Res.=2242024.03 | 38.1 | |
| IDEFICS-9BLLM=LLAMA-7B, Res.=2242024.03 | 25.2 | |
| InstructBLIPLLM Backbone=Vicuna-7B2023.12 | 23.7 | |
| InstructBLIPLLM=Vicuna-7B, Res.=2242024.03 | 23.7 | |
| LLaVALLM Backbone=Vicuna-7B2023.12 | 14.1 | |
| Qwen-VLLLM=Qwen-7B, Res.=4482024.03 | 7.4 |