AA Global-MMLU-Lite - Spanish — leaderboard
Metric: Accuracy (%). Source: artificialanalysis.ai. 118 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 94.42 |
| 2 | Gemini 3 Pro (Preview) (High) | 94.17 |
| 3 | Claude Opus 4.6 (Adaptive Reasoning, Max Effort) | 94 |
| 4 | Claude Opus 4.5 (Non-reasoning) | 93.83 |
| 5 | Claude Opus 4.5 (Thinking) | 93 |
| 6 | Grok 4 | 92.83 |
| 7 | GPT-5 (Medium) | 92.83 |
| 8 | GPT-5.1 (High) | 92.83 |
| 9 | GPT-5.1 Codex (High) | 92.75 |
| 10 | GPT-5 (High) | 92.67 |
| 11 | Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort) | 92.67 |
| 12 | Qwen 3.5 122B A10B (Reasoning) | 92.42 |
| 13 | Grok 4.20 0309 (Reasoning) | 92.17 |
| 14 | GPT-5.2 (Medium) | 92.08 |
| 15 | DeepSeek V3.2 Speciale | 91.92 |
Interactive version: theaggregate.ai/benchmark?slug=aa-global-mmlu-lite-spanish · How the rankings work · Data refreshed daily, snapshot 2026-07-22.