Neura Intelligence Index
Back to Leaderboard
MMMU-Pro
Harder version of MMMU with augmented candidate options and vision-only input.
multimodal0-100 percent
Model Rankings (64)
| # | Model | Score | Provider | Source |
|---|---|---|---|---|
| 1 | 83.6 | verified | ||
| 2 | 83.2 | OpenAI | verified | |
| 3 | 83 | OpenAI | verified | |
| 4 | B Seed 2.1 Pro🇨🇳 | 82.7 | ByteDance | verified |
| 5 | B Seed 2.1 Turbo🇨🇳 | 82.2 | ByteDance | verified |
| 6 | M Kimi K3🇨🇳 | 81.6 | Moonshot AI | verified |
| 7 | 81.2 | verified | ||
| 8 | 81.2 | OpenAI | verified | |
| 9 | 81 | verified | ||
| 10 | 80.7 | OpenAI | verified | |
| 11 | 80.5 | verified | ||
| 12 | M Muse Spark🇺🇸 | 80.4 | Meta | verified |
| 13 | M Kimi K2.6🇨🇳 | 80.1 | Moonshot AI | verified |
| 14 | 79.5 | OpenAI | verified | |
| 15 | Q Qwen3.7-Plus🇨🇳 | 79 | Alibaba Cloud / Qwen Team | verified |
| 16 | Q Qwen3.6 Plus🇨🇳 | 78.8 | Alibaba Cloud / Qwen Team | verified |
| 17 | M Kimi K2.5🇨🇳 | 78.5 | Moonshot AI | verified |
| 18 | 78.4 | OpenAI | verified | |
| 19 | 78.4 | OpenAI | verified | |
| 20 | 78.1 | MiniMax | verified | |
| 21 | X MiMo-V2.5🇨🇳 | 77.9 | Xiaomi | verified |
| 22 | 77.3 | Anthropic | verified | |
| 23 | 76.9 | verified | ||
| 24 | Q Qwen3.5-122B-A10B🇨🇳 | 76.9 | Alibaba Cloud / Qwen Team | verified |
| 25 | 76.8 | verified | ||
| 26 | 76.6 | OpenAI | verified | |
| 27 | 76.4 | OpenAI | verified | |
| 28 | 76 | OpenAI | verified | |
| 29 | Q Qwen3.6-27B🇨🇳 | 75.8 | Alibaba Cloud / Qwen Team | verified |
| 30 | 75.6 | Anthropic | verified | |
| 31 | Q Qwen3.6-35B-A3B🇨🇳 | 75.3 | Alibaba Cloud / Qwen Team | verified |
| 32 | Q Qwen3.5-35B-A3B🇨🇳 | 75.1 | Alibaba Cloud / Qwen Team | verified |
| 33 | Q Qwen3.5-27B🇨🇳 | 75 | Alibaba Cloud / Qwen Team | verified |
| 34 | 73.8 | verified | ||
| 35 | Q Qwen3 VL 235B A22B Thinking🇨🇳 | 69.3 | Alibaba Cloud / Qwen Team | verified |
| 36 | 69.1 | verified | ||
| 37 | Q Qwen3 VL 32B Thinking🇨🇳 | 68.1 | Alibaba Cloud / Qwen Team | verified |
| 38 | Q Qwen3 VL 235B A22B Instruct🇨🇳 | 68.1 | Alibaba Cloud / Qwen Team | verified |
| 39 | 66.1 | OpenAI | verified | |
| 40 | Q Qwen3 VL 32B Instruct🇨🇳 | 65.3 | Alibaba Cloud / Qwen Team | verified |
| 41 |  Nova 2 Pro🇺🇸 | 63.5 | Amazon | verified |
| 42 | Q Qwen3 VL 30B A3B Thinking🇨🇳 | 63 | Alibaba Cloud / Qwen Team | verified |
| 43 | 63 | Cohere | verified | |
| 44 | Nova 2 Lite🇺🇸 | 61.8 | Amazon | verified |
| 45 | Nova 2 Omni🇺🇸 | 61.4 | Amazon | verified |
| 46 | Q Qwen3 VL 8B Thinking🇨🇳 | 60.4 | Alibaba Cloud / Qwen Team | verified |
| 47 | Q Qwen3 VL 30B A3B Instruct🇨🇳 | 60.4 | Alibaba Cloud / Qwen Team | verified |
| 48 | 60 | Mistral AI | verified | |
| 49 | 59.9 | OpenAI | verified | |
| 50 | M Llama 4 Maverick🇺🇸 | 59.6 | Meta | verified |
| 51 | Q Qwen3 VL 4B Thinking🇨🇳 | 57 | Alibaba Cloud / Qwen Team | verified |
| 52 | Q Qwen3 VL 8B Instruct🇨🇳 | 55.9 | Alibaba Cloud / Qwen Team | verified |
| 53 | 54.3 | verified | ||
| 54 | Q Qwen3 VL 4B Instruct🇨🇳 | 53.2 | Alibaba Cloud / Qwen Team | verified |
| 55 | 52.6 | verified | ||
| 56 | Q Qwen2.5 VL 72B Instruct🇨🇳 | 51.1 | Alibaba Cloud / Qwen Team | verified |
| 57 | Q Qwen2.5 VL 32B Instruct🇨🇳 | 49.5 | Alibaba Cloud / Qwen Team | verified |
| 58 | Q Qwen2-VL-72B-Instruct🇨🇳 | 46.2 | Alibaba Cloud / Qwen Team | verified |
| 59 | M Llama 3.2 90B Instruct🇺🇸 | 45.2 | Meta | verified |
| 60 | 44.2 | verified | ||
| 61 | 38.5 | Microsoft | verified | |
| 62 | Q Qwen2.5 VL 7B Instruct🇨🇳 | 38.3 | Alibaba Cloud / Qwen Team | verified |
| 63 | Q Qwen2.5-Omni-7B🇨🇳 | 36.6 | Alibaba Cloud / Qwen Team | verified |
| 64 | M Llama 3.2 11B Instruct🇺🇸 | 33 | Meta | verified |