Multilingual Medical Reasoning on CUREMED-BENCH (test) benchmark leaderboard · SOTA2 Research

Multilingual Medical Reasoning on CUREMED-BENCH (test)

94.96ConsistencyCURE-MED-Qwen2.5-32B

Evaluation Results

Method	Links
CURE-MED-Qwen2.5-32B2026.01		94.96	70.04
CURE-MED-Qwen2.5-14B2026.01		90.27	63.74
HuatuoGPT-o1-70B2026.01		86.79	66.67
CURE-MED-Qwen2.5-7B2026.01		85.21	54.35
LLaMA-3.3-Instruct-70B2026.01		79.66	60.8
LLaMA-3.1-70B2026.01		75.68	54.65
CURE-MED-Qwen2.5-3B2026.01		74.28	42.93
Qwen2.5-Instruct-72B2026.01		70.73	58.8
OpenBioLLM-Llama3-70B2026.01		70.3	51.22
HuatuoGPT-o1-8B2026.01		67.3	46.86
CURE-MED-Qwen2.5-1.5B2026.01		57.6	28.32
UltraMedical LLaMA-3-8B2026.01		47.03	35.29
Ministral-8B2026.01		46.93	42.87
Qwen2.5-Instruct-32B2026.01		41.51	49.69
LLaMA-3.1-Instruct-8B2026.01		36.56	18.91
Qwen2.5-Instruct-14B2026.01		35.57	41.79
LLaMA-3-8B2026.01		31.58	28.93
MMed-LLaMA-3.1-70B2026.01		26.49	37.85
Apollo2-7B2026.01		25.63	15.93
Qwen2.5-Instruct-7B2026.01		25.44	29.56
LLaMA-3.2-3B2026.01		23.69	10.41
Gemma-9B2026.01		23.22	36.97
MMed-Llama-3-8B2026.01		21.38	28.09
Mistral-7B2026.01		18.7	15.23
Qwen2.5-Instruct-3B2026.01		8.39	10.83
BioMistral-7B2026.01		7.1	4.8
Qwen2.5-Instruct-1.5B2026.01		3.84	6.2
MedAlpaca-7B2026.01		3.5	2.47
OpenBioLLM-Llama3-8B2026.01		1.47	36.62
Meditron-7B2026.01		0.43	2.5
Gemma-7B2026.01		0.37	1.23
Meditron-70B2026.01		0.21	4.54
MedAlpaca-13B2026.01		0.1	0.07