PolyMATH

PolyMATH - Benchmark Leaderboard & Model Performance | AI Stats

Models Using This Benchmark

Organisation	Model	Reported	Top Score	Info	Self Reported	Source
Qwen	Qwen 3.6 Plus	01 Apr 2026	77.40%	-	Yes	Source
Qwen	Qwen 3.5 397B A17B	16 Feb 2026	73.30%	-	Yes	Source
Qwen	Qwen 3.5 27B	24 Feb 2026	71.20%	-	Yes	Source
Qwen	Qwen 3.5 Flash	23 Feb 2026	71.20%	inferred family alias from qwen3.5-27b (score=0.4147; benches=81)	Yes	Source
Qwen	Qwen 3.5 122B A10B	24 Feb 2026	68.90%	-	Yes	Source
Qwen	Qwen 3.5 35B A3B	24 Feb 2026	64.40%	-	Yes	Source
Qwen	Qwen 3 235B A22B Thinking 2507	-	60.10%	-	Yes	Source
Qwen	Qwen 3.5 9B	02 Mar 2026	57.30%	-	Yes	Source
Qwen	Qwen 3 Next 80B A3B Thinking	-	56.30%	-	Yes	Source
Qwen	Qwen 3 VL 32B Thinking	-	52%	-	Yes	-
Qwen	Qwen 3 VL 30B A3B Thinking	-	51.70%	-	Yes	-
Qwen	Qwen 3.5 4B	02 Mar 2026	51.10%	-	Yes	Source
Qwen	Qwen 3 8B	-	47.50%	inferred high-confidence family alias from qwen3-vl-8b-thinking (score=0.4600; benches=50)	Yes	-
Qwen	Qwen 3 Embedding 8B	-	47.50%	inferred family alias from qwen3-vl-8b-thinking (score=0.3850; benches=50)	Yes	-
Qwen	Qwen 3 Guard Gen 8B	-	47.50%	inferred family alias from qwen3-vl-8b-thinking (score=0.3400; benches=50)	Yes	-
Qwen	Qwen 3 Guard Stream 8B	-	47.50%	inferred family alias from qwen3-vl-8b-thinking (score=0.3371; benches=50)	Yes	-
Qwen	Qwen 3 Reranker 8B	-	47.50%	inferred family alias from qwen3-vl-8b-thinking (score=0.3850; benches=50)	Yes	-
Qwen	Qwen 3 VL 8B Thinking	-	47.50%	-	Yes	-
Qwen	Qwen 3 VL Embedding 8B	-	47.50%	inferred high-confidence family alias from qwen3-vl-8b-thinking (score=0.5232; benches=50)	Yes	-
Qwen	Qwen 3 VL Reranker 8B	-	47.50%	inferred high-confidence family alias from qwen3-vl-8b-thinking (score=0.5275; benches=50)	Yes	-
Qwen	Qwen 3 Next 80B A3B Instruct	-	45.90%	-	Yes	Source
Qwen	Qwen 3 4B Instruct 2507	-	44.60%	inferred family alias from qwen3-vl-4b-thinking (score=0.3462; benches=48)	Yes	-
Qwen	Qwen 3 4B SafeRL	-	44.60%	inferred family alias from qwen3-vl-4b-thinking (score=0.3850; benches=48)	Yes	-
Qwen	Qwen 3 Embedding 4B	-	44.60%	inferred family alias from qwen3-vl-4b-thinking (score=0.3850; benches=48)	Yes	-
Qwen	Qwen 3 Guard Gen 4B	-	44.60%	inferred family alias from qwen3-vl-4b-thinking (score=0.3400; benches=48)	Yes	-
Qwen	Qwen 3 4B Thinking 2507	-	44.60%	inferred family alias from qwen3-vl-4b-thinking (score=0.3462; benches=48)	Yes	-
Qwen	Qwen 3 Guard Stream 4B	-	44.60%	inferred family alias from qwen3-vl-4b-thinking (score=0.3371; benches=48)	Yes	-
Qwen	Qwen 3 Reranker 4B	-	44.60%	inferred family alias from qwen3-vl-4b-thinking (score=0.3850; benches=48)	Yes	-
Qwen	Qwen 3 VL 4B Thinking	-	44.60%	-	Yes	-
Qwen	Qwen 3 VL 30B A3B Instruct	-	44.30%	-	Yes	-
Qwen	Qwen 3 VL 32B Instruct	-	40.50%	-	Yes	-
Qwen	Qwen 3 VL 8B Instruct	-	30.40%	-	Yes	-
Qwen	Qwen 3 VL 4B Instruct	-	28.80%	-	Yes	-
Qwen	Qwen 3.5 2B	02 Mar 2026	26.10%	-	Yes	Source
Qwen	Qwen 3.5 0.8B	02 Mar 2026	8.20%	-	Yes	Source

Recorded Results

Average Score

Score Range

Leading Model

Models Using This Benchmark