Share your thoughts, 1 month free Claude Pro on usSee more

Multitask Language Understanding on MMLU Professional Medicine (test)

85.29Accuracy

Qwen3-32B

Updated 9d ago

Evaluation Results

Method	Links
Qwen3-32B 2026.04		85.29
Qwen3-8B-Base 2026.04		83.46
Qwen3-14B 2026.04		83.46
MedGemma-27B-it 2026.04		83.09
Qwen3-8B 2026.04		82.35
Qwen3-4B-Base 2026.04		78.31
Llama3.1-8B-Instruct 2026.04		76.47
Qwen3-4B 2026.04		76.1
Llama3.2-1B 2026.04		69.85
Llama3.1-8B 2026.04		69.85
Apollo-7B 2026.04		68.01
Mistral-7B-v0.3 2026.04		65.07
Qwen3-1.7B-Base 2026.04		64.71
Olmo3-7B 2026.04		63.6
Apertus-8B-Instruct 2026.04		62.87
Mistral-7B-Instruct-v0.3 2026.04		62.5
MedGemma-4B-it 2026.04		62.13
Apertus-8B 2026.04		60.29
EuroLLM-9B-Instruct 2026.04		59.56
GPT-OSS-20B 2026.04		59.19
Qwen3-1.7B 2026.04		58.09
SmolLM-3B 2026.04		55.88
Qwen3-0.6B-Base 2026.04		55.51
EuroLLM-9B 2026.04		55.51
BioMistral-7B 2026.04		55.15
Llama3.2-1B-Instruct 2026.04		51.84
Gaperon-8B 2026.04		48.53
Baguettotron 2026.04		44.85
Gaperon-1B 2026.04		44.85
Gemma3-270m-pt 2026.04		43.38
Qwen3-0.6B 2026.04		41.18
MedGemma-4B-pt 2026.04		38.24
Gemma3-270m-it 2026.04		36.76
Gemma3-1B-pt 2026.04		30.51
Gemma3-1B-it 2026.04		27.21