Benchmarks · knowledge
Global-MMLU-Lite
Global-MMLU-Lite: multilingual MMLU subset; accuracy.
44 models · 57 results · data as of 2026-11-06 · artificialanalysis.ai ↗
Leaderboard
Gemini 3.1 Pro Preview
All results
57 rows| # | Model | Effort | Global-MMLU-Lite |
|---|---|---|---|
| 1 | Gemini 3.1 Pro | 93.2% | |
| 2 | Gemini 3 Pro | high | 92.2% |
| 3 | Gemini 3 Flash | 91.4% | |
| 4 | Opus 4.5 | 91.3% | |
| 5 | GPT-5 | high | 90.7% |
| 6 | Opus 4.6 | high | 90.6% |
| 7 | Opus 4.5 | none | 90.6% |
| 8 | GPT-5.1 | high | 90.6% |
| 9 | GPT-5 | medium | 90.4% |
| 10 | Gemini 2.5 Pro | 90.3% | |
| 11 | Qwen 3.5 397B-A17B | 90.0% | |
| 12 | GPT-5.2 | medium | 89.8% |
| 13 | GPT-5.1 Codex | high | 89.7% |
| 14 | Grok 4 | 89.5% | |
| 15 | GPT-5.4 Mini | xhigh | 89.4% |
| 16 | Sonnet 4.5 | 89.3% | |
| 17 | Sonnet 4.6 | none | 89.2% |
| 18 | Qwen 3.5 122B-A10B | 89.1% | |
| 19 | Sonnet 4.6 | low | 88.2% |
| 20 | GPT-5 Mini | high | 87.4% |
| 21 | GPT-5.2 | none | 86.8% |
| 22 | Qwen 3.5 27B | 86.8% | |
| 23 | DeepSeek V3.2 | 86.7% | |
| 24 | Qwen 3.5 397B-A17B | none | 86.5% |
| 25 | Qwen 3.5 35B-A3B | 86.4% | |
| 26 | Qwen 3.5 122B-A10B | none | 85.9% |
| 27 | GLM 4.6 | 85.6% | |
| 28 | GPT-5.1 Codex Mini | high | 85.0% |
| 29 | MiMo V2 Flash | 84.4% | |
| 30 | GPT-5 | minimal | 84.4% |