View overall rankings across multimodal AI models capable of reasoning over visual inputs.
Lab Rank | Model Score | Rank Spread | ||
|---|---|---|---|---|
| 1 | Anthropic claude-fable-5-high · Proprietary | 1309±7 | 1 | 110 |
| 2 | Alibaba qwen3.8-max · Proprietary | 1301±7 | 2 | 116 |
| 3 | Google gemini-3.7-flash-high · Proprietary | 1296±10 | 6 | 128 |
| 4 | Meta muse-spark · Proprietary | 1294±9 | 8 | 132 |
| 5 | OpenAI gpt-6.1-sol-max · Proprietary | 1291±16 | 10 | 137 |
| 6 | SpaceXAI grok-4.5 · Proprietary | 1279±7 | 28 | 741 |
| 7 | Z.ai glm-5.3-flash · MIT | 1278±9 | 30 | 742 |
| 8 | Moonshot kimi-k2.6 · Modified MIT | 1265±7 | 38 | 2850 |
| 9 | StepFun Step 5 Preview · Proprietary | 1260±24 | 45 | 968 |
| 10 | Bytedance dola-seed-2.0-pro · Proprietary | 1257±7 | 46 | 3556 |
| 11 | Xiaomi mimo-v2.6-flash · MIT | 1247±13 | 53 | 3569 |
| 12 | MiniMax minimax-m3 · MiniMax Community License | 1237±7 | 63 | 5172 |
| 13 | Baidu ernie-5.0-preview-1220 · Proprietary | 1219±11 | 72 | 6287 |
| 14 | Thinky Inkling Small · Apache 2.0 | 1206±9 | 81 | 7193 |
| 15 | Mistral mistral-large-3 · Apache 2.0 | 1201±8 | 85 | 7595 |
| 16 | Tencent hunyuan-vision-1.5-thinking · Proprietary | 1156±12 | 108 | 99118 |
| 17 | Ai2 molmo-2-8b · Apache 2.0 | 1112±20 | 128 | 117133 |
| 18 | internvl2-26b · MIT | 1025±13 | 143 | 138149 |
| 19 | Amazon amazon-nova-lite-v1.0 · Proprietary | 1019±15 | 144 | 138152 |
| 20 | Cohere c4ai-aya-vision-32b · CC-BY-NC-4.0 | 1000±22 | 148 | 140155 |
| 21 | Nvidia nvila-internal-15b-v1 · - | 985±20 | 152 | 144157 |
| 22 | Microsoft phi-3.5-vision-instruct · MIT | 919±16 | 158 | 158158 |