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

Radiology Question Answering on RadQA and Board-RadQA Combined (test)

91Accuracy

Gemini-3.1-Pro

Updated 1mo ago

Evaluation Results

Method
Gemini-3.1-Pro 2026.03	91	154	0.012
DeepSeek-V3 2026.03	89	150	0.218
o3 2026.03	89	150	0.239
Claude-Sonnet-4.6 2026.03	88	148	0.577
DeepSeek-R1 2026.03	88	148	-
MiMo-V2-Flash 2026.03	87	147	0.063
GPT-5 2026.03	87	147	0.402
GPT-5.2 2026.03	87	147	0.035
Palmyra-X5 2026.03	86	146	-
Palmyra-X5 2026.03	86	145	1
Llama3.3-70B 2026.03	86	145	0.144
Mistral Large (123B) 2026.03	86	146	0.166
Qwen3-235B 2026.03	86	146	-
Claude-Sonnet-4.6 2026.03	85	144	-
GLM-5 2026.03	85	144	0.527
Kimi-K2.5 2026.03	85	144	0.729
Llama4-Scout-16E 2026.03	85	143	0.585
o3 2026.03	85	144	-
Qwen3-235B 2026.03	85	144	0.877
DeepSeek R1-70B 2026.03	84	142	-
DeepSeek R1-70B 2026.03	84	142	1
DeepSeek-R1 2026.03	84	142	0.343
DeepSeek-V3 2026.03	84	142	-
GPT-5 2026.03	84	142	-
MedGemma-27B-text-it 2026.03	84	142	0.22
Qwen2.5-70B 2026.03	84	142	0.229
Kimi-K2.5 2026.03	83	141	-
Gemma-3-27B-it 2026.03	83	141	0.009
Gemini-3.1-Pro 2026.03	82	138	-
GLM-5 2026.03	82	139	-
Llama4-Scout-16E 2026.03	82	139	-
GPT-4-turbo 2026.03	82	138	0.277
Mistral Large (123B) 2026.03	81	137	-
Qwen3-8B 2026.03	81	137	0.218
MiniMax-M2.5 2026.03	80	135	0.999
Llama3.3-70B 2026.03	80	135	-
MiniMax-M2.5 2026.03	79	134	-
MiMo-V2-Flash 2026.03	79	133	-
Llama3-Med42-70B 2026.03	79	134	0.417
GPT-5.2 2026.03	79	134	-
MedGemma-27B-text-it 2026.03	79	133	-
Qwen2.5-14B 2026.03	79	134	0.354
Qwen2.5-70B 2026.03	79	134	-
Qwen2.5-7B 2026.03	78	132	0.002
GPT-3.5-turbo 2026.03	77	130	0.012
GPT-4-turbo 2026.03	77	130	-
LFM2.5-1.2B-Thinking 2026.03	76	129	0.001
Ministral-8B 2026.03	76	128	0.001
Llama3.3-8B 2026.03	75	126	0.275
Llama3-Med42-70B 2026.03	75	127	-
MedGemma-4B-it 2026.03	75	127	0.024
Qwen3-8B 2026.03	75	127	-
Llama3-Med42-8B 2026.03	74	125	0.024
Qwen2.5-14B 2026.03	74	126	-
Qwen2.5-3B 2026.03	73	124	0.038
Gemma-3-4B-it 2026.03	71	121	0.001
Gemma-3-27B-it 2026.03	71	120	-
Llama3.3-8B 2026.03	69	117	-
GPT-3.5-turbo 2026.03	64	108	-
MedGemma-4B-it 2026.03	64	108	-
Llama3-Med42-8B 2026.03	63	107	-
Qwen2.5-3B 2026.03	63	107	-
Qwen2.5-7B 2026.03	63	107	-
LFM2.5-1.2B-Thinking 2026.03	56	94	-
Ministral-8B 2026.03	54	91	-
Gemma-3-4B-it 2026.03	51	86	-
Qwen2.5-0.5B 2026.03	48	81	0.079
Qwen2.5-0.5B 2026.03	37	63	-