# Artificial Analysis Global-MMLU-Lite (AA Global-MMLU-Lite) > Global MMLU Lite is a multilingual version of MMLU with 400 questions per language, half culturally sensitive and half culturally agnostic. This page holds two independent runs: Artificial Analysis's, in the main table, and Kaggle's. Canonical page: https://benchlm.ai/benchmarks/global-mmlu-lite - Category: [Multilingual](/multilingual) - Last updated: November 11, 2026 ## About AA Global-MMLU-Lite - Year: 2026 - Tasks: Multilingual knowledge questions - Format: Accuracy - Difficulty: Multilingual professional knowledge - Paper: [Artificial Analysis Global-MMLU-Lite Benchmark Leaderboard](https://artificialanalysis.ai/evaluations/global-mmlu-lite) The main table is the average across languages that Artificial Analysis publishes, stored as a display-only multilingual score. Kaggle's run of the same test, an average of 16 language dimensions, is shown in its own table with a comparison of the two. AA Global-MMLU-Lite is currently displayed on BenchLM for reference, but it is excluded from the weighted scoring formula. ## Leaderboard (14 models) | Rank | Model | Creator | Score | |------|-------|---------|-------| | 1 | [Gemini 3.1 Pro](/models/gemini-3-1-pro) | Google | 93.2% | | 2 | [Gemini 3 Flash](/models/gemini-3-flash) | Google | 92.7% | | 3 | [Gemini 3 Pro](/models/gemini-3-pro) | Google | 92.2% | | 4 | [Claude Opus 4.6 (Adaptive)](/models/claude-opus-4-6-thinking) | Anthropic | 92.2% | | 5 | [Claude Opus 4.5 Thinking](/models/claude-opus-4-5-thinking) | Anthropic | 91.3% | | 6 | [GPT-5 (high)](/models/gpt-5-high) | OpenAI | 90.7% | | 7 | [GPT-5.1](/models/gpt-5-1) | OpenAI | 90.6% | | 8 | [Gemini 2.5 Pro](/models/gemini-2-5-pro) | Google | 90.3% | | 9 | [Grok 4](/models/grok-4) | xAI | 89.5% | | 10 | [GPT-5.4 mini](/models/gpt-5-4-mini) | OpenAI | 89.4% | | 11 | [Qwen3.5-122B-A10B](/models/qwen3-5-122b-a10b) | Alibaba | 89.1% | | 12 | [Qwen3.5-27B](/models/qwen3-5-27b) | Alibaba | 86.8% | | 13 | [Qwen3.5-35B-A3B](/models/qwen3-5-35b-a3b) | Alibaba | 86.4% | | 14 | [DeepSeek-R1](/models/deepseek-r1) | DeepSeek | 86.0% | ## Global MMLU Lite (Kaggle) Evaluation results published by [Cohere Labs](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite). Source snapshot: Source retrieved 2026-10-20T15:19:04.746Z. Published proportions are displayed as percentages. Model configurations, evaluation dates, task versions, and source uncertainty remain attached to the snapshot. Missing results are not zero. - [Kaggle leaderboard and Apache 2.0 declaration](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) - [Cohere Labs Global-MMLU-Lite dataset card](https://huggingface.co/datasets/CohereLabs/Global-MMLU-Lite) - [Global-MMLU paper](https://arxiv.org/abs/2412.03304) | Source model | Configuration | Kaggle average | |---|---|---| | [Claude Opus 5](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 95.7% | | [Gemini 3.5 Flash](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 95.4% | | [Gemini 4 Argon](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 95.4% | | [Claude Opus 5.5](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 95.4% | | [Gemini 3.8 Flash](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 95.3% | | [Gemini 3.7 Flash](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 94.9% | | [Claude Opus 4.8](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 94.9% | | [Gemini 3 Pro Preview](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 94.5% | | [Claude Opus 4.1](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 94.3% | | [Gemini 3.6 Flash](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 94.3% | | [GPT-5.6 Sol](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 94.1% | | [GPT-6 Sol](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 94.1% | | [Grok 4.5](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 93.3% | | [Gemini 2.5 Pro](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 93.2% | | [GPT-5.6 Terra](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 92.7% | | [Claude Sonnet 5.5](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 92.6% | | [GPT-6 Astra](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 92.6% | | [GPT-6 Luna](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 92.0% | | [Gemini 2.5 Flash](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 91.5% | | [Gemini 2.5 Flash Preview](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 90.9% | | [Gemma 4 31B](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 90.7% | | [Claude Sonnet 4](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 90.6% | | [Gemini 3.5 Flash-Lite](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 89.5% | | [GPT-5](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 89.0% | | [Grok 4](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 88.8% | | [Qwen 3 235B A22B Instruct 2506](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 88.0% | | [GPT-4.1](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 87.5% | | [Claude Haiku 5.5](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 87.3% | | [o4 mini](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 87.0% | | [Gemma 4 26B A4B](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 85.7% | | [Command A](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 83.9% | | [GPT-5.6 Luna](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 83.1% | | [Claude 3.7 Sonnet](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 80.8% | | [deepseek-v3.1](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 80.4% | | [GPT-6.1 Sol](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 78.9% | | [Mistral Small 3.1](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 78.5% | | [o3 mini](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 78.0% | | [Gemma 3 27B](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 76.3% | | [Granite 4.0 Small](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 75.0% | | [Aya Expanse 32B](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 73.5% | | [DeepSeek-R1](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 67.4% | | [Grok 3 Mini](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 67.3% | | [Gemma 3 4B](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 65.1% | | [Claude 3.5 Haiku](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 61.1% | | [Mistral Medium 3](https://www.kaggle.com/benchmarks/cohere-labs/global-mmlu-lite) | Not reported | 55.1% | ## FAQ ### What does AA Global-MMLU-Lite measure? Global MMLU Lite is a multilingual version of MMLU with 400 questions per language, half culturally sensitive and half culturally agnostic. This page holds two independent runs: Artificial Analysis's, in the main table, and Kaggle's. ### Which model scores highest on AA Global-MMLU-Lite? Gemini 3.1 Pro by Google currently leads with a score of 93.2% on AA Global-MMLU-Lite. ### How many models are evaluated on AA Global-MMLU-Lite? 14 AI models have published results on AA Global-MMLU-Lite in the BenchLM catalog. ### Does AA Global-MMLU-Lite affect BenchLM's overall score? Not directly. AA Global-MMLU-Lite is still displayed on BenchLM for reference, but it is excluded from the weighted scoring formula. ## Compare Top Models on AA Global-MMLU-Lite - [Gemini 3.1 Pro vs Gemini 3 Flash](/compare/gemini-3-1-pro-vs-gemini-3-flash) - [Gemini 3 Flash vs Gemini 3 Pro](/compare/gemini-3-flash-vs-gemini-3-pro) - [Gemini 3 Pro vs Claude Opus 4.6 (Adaptive)](/compare/claude-opus-4-6-thinking-vs-gemini-3-pro) - [Claude Opus 4.6 (Adaptive) vs Claude Opus 4.5 Thinking](/compare/claude-opus-4-5-thinking-vs-claude-opus-4-6-thinking)