Benchmarks · knowledge
MMMU-Pro
MMMU-Pro: multimodal college-level understanding; accuracy.
105 models · 207 results · data as of 2026-11-06 · artificialanalysis.ai ↗
Leaderboard
Gemini 4 Carbon medium
More from this benchmark
3 published measuresScore vs any published measure3 measures: sub-scores, splits, counts
76 more models under All models in settings.
Breakdown
darker is better within each column · click a column to sort| Model | |||
|---|---|---|---|
| Gemini 4 Carbon medium | 2,284,674 | 554,565 | 2,043,369 |
| Opus 5.5 max | 1,210,518 | 630,042 | 8,407,797 |
| GPT-6 Astra max | 1,000,367 | 141,040 | 1,898,414 |
| GPT-6.1 Sol max | 1,000,367 | 135,231 | 1,853,389 |
| Gemini 3.8 Flash high | 2,246,200 | 444,646 | 7,849,823 |
| Gemini 3.7 Flash high | 2,246,200 | 412,563 | 2,635,088 |
| Opus 5 max | 1,207,066 | 505,657 | 4,292,435 |
| GPT-6.1 Terra max | 1,000,367 | 97,312 | 4,793,870 |
| Gemini 3.5 Flash high | 2,246,200 | 561,223 | 4,088,031 |
| Kimi K3.1 max | 1,210,646 | 171,062 | 7,484,413 |
| GPT-5.6 Sol max | 999,093 | 121,881 | 2,978,298 |
| Gemini 3.6 Flash high | 2,215,607 | 480,256 | 3,688,917 |
| GPT-6 Sol max | 1,000,367 | 76,832 | 2,667,643 |
| Gemini 3.1 Pro | 2,238,058 | 728,024 | 10,295,861 |
| Qwen 3.8 Max | 898,118 | 368,846 | 4,501,114 |
| Muse Spark 1.3 xhigh | 1,027,117 | 251,133 | 8,585,609 |
| Inkling 2 Small | 759,096 | 1,228,852 | 12,073,824 |
| GPT-5.5 medium | 1,013,956 | 165,999 | 1,348,496 |
| Muse Spark Open xhigh | 1,025,901 | 257,336 | 7,321,813 |
| GPT-5.6 Terra max | 1,590,767 | 124,781 | 7,480,205 |
| DeepSeek V4.1 Flash 1017 max | 834,480 | 119,718 | 13,335,560 |
| Muse Spark | 1,495,382 | 272,488 | 4,631,129 |
| Kimi K3 max | 1,210,646 | 171,047 | 9,595,246 |
| Qwen 3.7 Plus | 897,709 | 846,301 | 7,376,483 |
| Grok 4.5 high | 952,142 | 433,862 | 7,415,809 |
All results
207 rows| # | Model | Effort | MMMU-Pro |
|---|---|---|---|
| 1 | Gemini 4 Carbon | medium | 89.2% |
| 2 | Opus 5.5 | max | 87.7% |
| 3 | GPT-6 Astra | max | 86.9% |
| 4 | Opus 5.5 | xhigh | 86.6% |
| 5 | GPT-6 Astra | high | 86.4% |
| 6 | GPT-6 Astra | xhigh | 86.2% |
| 7 | GPT-6.1 Sol | max | 86.0% |
| 8 | Opus 5.5 | high | 85.8% |
| 9 | Opus 5.5 | medium | 85.7% |
| 10 | Gemini 3.8 Flash | high | 85.6% |
| 11 | Gemini 3.7 Flash | high | 85.5% |
| 12 | GPT-6 Astra | medium | 85.1% |
| 13 | GPT-6.1 Sol | xhigh | 85.1% |
| 14 | GPT-6.1 Sol | high | 84.9% |
| 15 | Gemini 3.7 Flash | low | 84.9% |
| 16 | Opus 5 | max | 84.7% |
| 17 | Gemini 3.7 Flash | medium | 84.7% |
| 18 | Opus 5.5 | low | 84.7% |
| 19 | GPT-6.1 Terra | max | 84.7% |
| 20 | GPT-6 Astra | low | 84.6% |
| 21 | Gemini 3.8 Flash | low | 84.5% |
| 22 | Gemini 3.5 Flash | high | 84.3% |
| 23 | Gemini 3.8 Flash | medium | 84.2% |
| 24 | Opus 5 | xhigh | 84.0% |
| 25 | GPT-6.1 Sol | medium | 83.9% |
| 26 | Gemini 3.5 Flash | medium | 83.9% |
| 27 | Kimi K3.1 | max | 83.9% |
| 28 | GPT-6.1 Terra | high | 83.8% |
| 29 | GPT-6.1 Terra | xhigh | 83.8% |
| 30 | GPT-5.6 Sol | max | 83.4% |