Multimodal Understanding on MM-Bench CN (test)
79.2AccuracyCambrian-34B
Evaluation results
| Method | Links | |
|---|---|---|
| Cambrian-34BSize=34B2024.08 | 79.2 | |
| LLaVA-NeXT-Yi-34BSize=34B2024.08 | 79 | |
| GPT-4V (2023.11.06)2024.08 | 74.4 | |
| Gemini Pro2024.08 | 74.3 | |
| MiniCPM-Llama3-V 2.5Size=8.5B2024.08 | 74.2 | |
| Bunny-Llama-3-8BSize=8.4B2024.08 | 73.9 | |
| DeepSeek-VL-7BSize=7.3B2024.08 | 71.4 | |
| Yi-VL-34BSize=34B2024.08 | 70.7 | |
| CogVLM2-19B-ChatSize=19B2024.08 | 69.8 | |
| Idefics2Size=8.0B2024.08 | 68.6 | |
| Cambrian-8BSize=8.3B2024.08 | 67.9 | |
| Yi-VL-6BSize=6.7B2024.08 | 66.6 | |
| MiniCPM-V 2.0Size=2.8B2024.08 | 66.5 | |
| XTuner-Llama-3-8B-v1.1Size=8.4B2024.08 | 63.2 | |
| DeepSeek-VL-1.3BSize=1.7B2024.08 | 62.9 | |
| MiniCPM-V 1.0Size=2.8B2024.08 | 62.6 | |
| Claude 3 Opus2024.08 | 59.2 | |
| Qwen-VL-ChatSize=9.6B2024.08 | 56.3 | |
| CogVLM-ChatSize=17B2024.08 | 55.9 |