Skip to main content
Radar

Every change to the models you run, with its source and its date. Releases, price changes, retirements, API changes, and incidents.Every change to the models you run, with its source.

Follow model changes
BenchLM

MMLU-ProX

Data verified 36 confirmed releases in the last 30 daysFollow model changes

A multilingual extension of professional-level academic evaluation across many languages.

Top models on MMLU-ProX — September 18, 2026

As of September 18, 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%).

1Proprietary

Qwen3.7 Max

Alibaba

87%
Overall 66.87Context 1M
2Proprietary

Claude Opus 4.5

Anthropic

85.7%
Overall 58.42Context 200K
3Proprietary

Qwen3.7 Plus

Alibaba

85.4%
Overall 61.6Context 1M

12 modelsMultilingual100% of category scoreCurrentUpdated September 18, 2026

Leaderboard (12 models)

Score
1
Qwen3.7 MaxAlibaba · Closed
87%
2
Claude Opus 4.5Anthropic · Closed
85.7%
3
Qwen3.7 PlusAlibaba · Closed
85.4%
4
Qwen3.6 PlusAlibaba · Closed
84.7%
5
Qwen3.5 397BAlibaba · Open weight
84.7%
6
GLM-5Z.AI · Open weight
83.1%
7
Nemotron 3 UltraNVIDIA · Open weight
83%
8
Kimi K2.5Moonshot AI · Open weight
82.3%
9
Qwen3.5-27BAlibaba · Open weight
82.2%
10
Qwen3.5-122B-A10BAlibaba · Open weight
82.2%
11
Qwen3.5-35B-A3BAlibaba · Open weight
81%
12
Qwen3 235B 2507Alibaba · Open weight
79.4%

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.

12 models have been evaluated on MMLU-ProX. The benchmark falls in the Multilingual category. This category carries a 7% weight in BenchLM.ai's overall scoring system. Within that category, MMLU-ProX contributes 100% of the category score, so strong performance here directly affects a model's 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.

MMLU-ProX: A Multilingual Benchmark for Advanced Large Language Model Evaluation

Freshness and provenance

Version

MMLU-ProX 2025

Refresh cadence

Static

Staleness state

Current

Question availability

Public benchmark set

Current

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?

12 AI models have been evaluated on MMLU-ProX on BenchLM.

Compare top models on MMLU-ProX

Qwen3.7 Max vs Claude Opus 4.5Claude Opus 4.5 vs Qwen3.7 PlusQwen3.7 Plus vs Qwen3.6 PlusQwen3.6 Plus vs Qwen3.5 397B
Last updated: September 18, 2026 · BenchLM version MMLU-ProX 2025

Know when it’s worth switching models

The model to choose, the cheaper alternative, and the release we would wait on.

Read a sample issue

Join 2,000+ readers.

One email each week. Unsubscribe anytime.