Skip to content

The Aggregate

Unified LLM rankings, daily

Models What's New Trends Benchmarks Skill Maps
How It Works Guesswork Metrics
The Aggregate
Loading data...

AA Global-MMLU-Lite - German — leaderboard

Metric: Accuracy (%). Source: artificialanalysis.ai. 119 models tracked.

Top models

#ModelScore
1Gemini 3.1 Pro (Preview)94.75
2Claude Opus 4.6 (Adaptive Reasoning, Max Effort)93.42
3Gemini 3 Pro (Preview) (High)93.17
4Claude Opus 4.5 (Thinking)93
5GPT-5 (High)92.5
6Gemini 2.5 Pro92.42
7GPT-5 (Medium)92.33
8Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort)92.25
9Claude Opus 4.5 (Non-reasoning)92.25
10GPT-5.1 (High)91.83
11Grok 4.20 0309 (Reasoning)91.75
12Claude Sonnet 4.5 (Thinking)91.67
13Qwen 3.5 122B A10B (Reasoning)91.58
14Grok 491.5
15GPT-5.2 (Medium)91.5

Interactive version: theaggregate.ai/benchmark?slug=aa-global-mmlu-lite-german · How the rankings work · Data refreshed daily, snapshot 2026-07-22.

Built by Mikhail Doroshenko — AI researcher, co-author of Humanity’s Last Exam. Independent project; no affiliation with any AI lab.