MMLU-ProX
A multilingual extension of professional-level academic evaluation across many languages.
Top models on MMLU-ProX — November 4, 2026
As of November 4, 2026, Qwen3.7 Max leads the MMLU-ProX leaderboard with 87% , followed by Claude Opus 4.5 (85.7%) and Qwen3.7 Plus (85.4%).
Qwen3.7 Max
Alibaba
Claude Opus 4.5
Anthropic
Qwen3.7 Plus
Alibaba
17 modelsMultilingual100% of category scoreCurrentUpdated November 4, 2026
Leaderboard (17 models)
| Rank | Model / configuration | Score | Parameters (B) | Open / closed |
|---|---|---|---|---|
| 1 | Qwen3.7 MaxAlibaba | 87% | Not reported | Closed |
| 2 | Claude Opus 4.5Anthropic | 85.7% | Not reported | Closed |
| 3 | Qwen3.7 PlusAlibaba | 85.4% | Not reported | Closed |
| 4 | Qwen3.6 PlusAlibaba | 84.7% | Not reported | Closed |
| 5 | Qwen3.5 397BAlibaba | 84.7% | Not reported | Open |
| 6 | GLM-5Z.AI | 83.1% | Not reported | Open |
| 7 | Nemotron 3 UltraNVIDIA | 83% | Not reported | Open |
| 8 | Kimi K2.5Moonshot AI | 82.3% | Not reported | Open |
| 9 | Qwen3.5-122B-A10BAlibaba | 82.2% | Not reported | Open |
| 10 | Qwen3.5-27BAlibaba | 82.2% | Not reported | Open |
| 11 | Qwen3.5-35B-A3BAlibaba | 81% | Not reported | Open |
| 12 | Qwen3 235B 2507 (Reasoning)Alibaba | 81% | Not reported | Open |
| 13 | Qwen3 235B 2507Alibaba | 79.4% | Not reported | Open |
| 14 | GPT-4.1OpenAI | 72.7% | Not reported | Closed |
| 15 | DeepSeek V3 0324DeepSeek | 70.5% | Not reported | Open |
| 16 | GPT-4oOpenAI | 61.1% | Not reported | Closed |
| 17 | Phi-4Microsoft | 49.9% | Not reported | Open |
According to BenchLM.ai, Qwen3.7 Max leads the MMLU-ProX benchmark with a score of 87%, followed by Claude Opus 4.5 (85.7%) and Qwen3.7 Plus (85.4%). The top models are clustered within 1.6 points, suggesting this benchmark is nearing saturation for frontier models.
17 models have been evaluated on MMLU-ProX. The benchmark falls in the Multilingual category. The Multilingual leaderboard ranks models by a weighted category score, and MMLU-ProX contributes 100% of it. BenchAlign v5.8 also uses it in the overall ranking.
About MMLU-ProX
Year
2025
Tasks
Multilingual professional QA
Format
Multilingual multiple choice
Difficulty
Professional multilingual
MMLU-ProX expands multilingual evaluation beyond translated arithmetic, making it a better signal for broad cross-lingual reasoning and knowledge.
Freshness and provenance
Version
MMLU-ProX 2025
Refresh cadence
Static
Staleness state
Current
Question availability
Public benchmark set
BenchLM uses freshness metadata to decide whether a benchmark should still be treated as a strong differentiator, a benchmark to watch, or a display-only reference. For the full scoring policy, see the BenchLM methodology page.
Questions
What does MMLU-ProX measure?
A multilingual extension of professional-level academic evaluation across many languages.
Which model scores highest on MMLU-ProX?
Qwen3.7 Max by Alibaba currently leads with a score of 87% on MMLU-ProX.
How many models are evaluated on MMLU-ProX?
17 AI models have published results on MMLU-ProX in the BenchLM catalog.
Compare top models on MMLU-ProX
Know when it’s worth switching models
The model to choose, the cheaper alternative, and the release we would wait on.
Read a sample issueJoin 6,100+ subscribers.
One email each week. Unsubscribe anytime.