MMLU-ProX

AI benchmark results and model performance

Type: percentageGeneral

Recorded Results

Average Score

71.15%

Score Range

34.60% - 84.70%

Leading Model (lowest score)

34.60% - Qwen 3.5 0.8B

Scores Over Time

Individual benchmark scores plotted by date.

Models Using This Benchmark

Organisation	Model	Reported	Top Score	Info	Self Reported	Source
Qwen	Qwen 3.5 0.8B	02 Mar 2026	34.60%	-	Yes	Source
Qwen	Qwen 3.5 2B	02 Mar 2026	52.30%	-	Yes	Source
Qwen	Qwen 3 VL 4B Instruct	-	59.40%	-	Yes	-
Nvidia	Nemotron 3 Nano 30B A3B	15 Dec 2025	59.50%	-	Yes	Source
Qwen	Qwen 3 4B Instruct 2507	-	65%	inferred family alias from qwen3-vl-4b-thinking (score=0.3462; benches=48)	Yes	-
Qwen	Qwen 3 4B SafeRL	-	65%	inferred family alias from qwen3-vl-4b-thinking (score=0.3850; benches=48)	Yes	-
Qwen	Qwen 3 4B Thinking 2507	-	65%	inferred family alias from qwen3-vl-4b-thinking (score=0.3462; benches=48)	Yes	-
Qwen	Qwen 3 Embedding 4B	-	65%	inferred family alias from qwen3-vl-4b-thinking (score=0.3850; benches=48)	Yes	-
Qwen	Qwen 3 Guard Gen 4B	-	65%	inferred family alias from qwen3-vl-4b-thinking (score=0.3400; benches=48)	Yes	-
Qwen	Qwen 3 Guard Stream 4B	-	65%	inferred family alias from qwen3-vl-4b-thinking (score=0.3371; benches=48)	Yes	-
Qwen	Qwen 3 Reranker 4B	-	65%	inferred family alias from qwen3-vl-4b-thinking (score=0.3850; benches=48)	Yes	-
Qwen	Qwen 3 VL 4B Thinking	-	65%	-	Yes	-
Qwen	Qwen 3 VL 8B Instruct	-	65.40%	-	Yes	-
Qwen	Qwen 3 8B	-	70.70%	inferred high-confidence family alias from qwen3-vl-8b-thinking (score=0.4600; benches=50)	Yes	-
Qwen	Qwen 3 Guard Stream 8B	-	70.70%	inferred family alias from qwen3-vl-8b-thinking (score=0.3371; benches=50)	Yes	-
Qwen	Qwen 3 VL 8B Thinking	-	70.70%	-	Yes	-
Qwen	Qwen 3 VL Embedding 8B	-	70.70%	inferred high-confidence family alias from qwen3-vl-8b-thinking (score=0.5232; benches=50)	Yes	-
Qwen	Qwen 3 VL Reranker 8B	-	70.70%	inferred high-confidence family alias from qwen3-vl-8b-thinking (score=0.5275; benches=50)	Yes	-
Qwen	Qwen 3 VL 30B A3B Instruct	-	70.90%	-	Yes	-
Qwen	Qwen 3.5 4B	02 Mar 2026	71.50%	-	Yes	Source
Qwen	Qwen 3 VL 32B Instruct	-	73.40%	-	Yes	-
Qwen	Qwen 3 VL 30B A3B Thinking	-	76.10%	-	Yes	-
Qwen	Qwen 3.5 9B	02 Mar 2026	76.30%	-	Yes	Source
Qwen	Qwen 3 Next 80B A3B Instruct	-	76.70%	MMLU-ProX	Yes	Source
Qwen	Qwen 3 VL 32B Thinking	-	77.20%	-	Yes	-
Qwen	Qwen 3 VL 235B A22B Instruct	-	77.80%	-	Yes	-
Qwen	Qwen 3 Next 80B A3B Thinking	-	78.70%	MMLU-ProX	Yes	Source
Nvidia	Nemotron 3 Super	11 Mar 2026	79.36%	-	Yes	Source
Qwen	Qwen 3 VL 235B A22B Thinking	-	80.60%	-	Yes	-
Qwen	Qwen 3 235B A22B Thinking 2507	-	81%	MMLU-ProX	Yes	Source
Qwen	Qwen 3.5 35B A3B	24 Feb 2026	81%	-	Yes	Source
Qwen	Qwen 3.5 122B A10B	24 Feb 2026	82.20%	-	Yes	Source
Qwen	Qwen 3.5 27B	24 Feb 2026	82.20%	-	Yes	Source
Qwen	Qwen 3.5 Flash	23 Feb 2026	82.20%	inferred family alias from qwen3.5-27b (score=0.4147; benches=81)	Yes	Source
Qwen	Qwen 3.5 397B A17B	16 Feb 2026	84.70%	-	Yes	Source
Qwen	Qwen 3.6 Plus	01 Apr 2026	84.70%	-	Yes	Source