Clinical Visual Question Answering on GMAI-MMBench standard MCQA benchmark leaderboard · SOTA2 Research

Clinical Visual Question Answering on GMAI-MMBench standard MCQA

47AccuracyQwen2.5-VL-72B

Evaluation Results

Method	Links
Qwen2.5-VL-72B2026.06		47	72	66.3	67.5	65.7	65.8	66.6	61.2	61.2	20.5	31.2	30.2	28.1	26.4	19.7	26.1	23.1
Qwen2.5-VL-32B2026.06		44.1	67	59.3	58.8	61.5	61.8	59.8	60	59.5	22.6	19.9	20.3	32.8	30.3	17.6	23	16.9
GPT-4.1-Mini2026.06		43	64.2	62.8	63.3	57.1	57.2	62.9	57	56.8	39.8	43.2	39.4	42.6	39.6	41.5	41.8	39.7
LLaVA-OV-8B2026.06		42.8	63.6	60.4	61.4	61.7	62	60.4	58.4	58.3	20.3	43.7	32.4	27.9	25	15	20	19.8
Qwen2-VL-7B2026.06		42.4	65.6	64.6	65.2	63.9	63.7	65.7	63.2	63.6	12.6	13.5	12.9	26.9	28.7	12.5	26.8	26.6
Qwen2.5-VL-7B2026.06		42.2	63.6	60	61.9	60.9	60.9	59.9	59.1	58.9	21.1	25	19.2	34.1	32.8	16.9	24.3	23.2
MedGemma-27B2026.06		39.3	55.1	54.2	54.6	56.1	56.2	53.9	53.4	53.6	53.1	18.7	27.3	39	35.3	19	32.3	29.8
LLaVA-v1.6-34B2026.06		39	67	62.3	64.6	62.8	63.2	62.4	59.1	58.8	21.2	48.8	35.9	30.8	28.1	18.6	24.3	23.1
MedGemma-4B2026.06		37.2	57.8	53.9	53.3	50.8	50.8	53.9	51.6	51.6	55.9	46.7	43.9	59.4	59.2	44.9	52.2	52.1
LLaVA-v1.6-7B2026.06		35.5	62.1	60.2	61	59.5	59.6	60.2	59.4	59.3	22.2	27.5	20.5	32.9	30.9	20.9	30.9	26.7
Molmo-72B2026.06		35.2	59.8	51.3	49.7	55.7	55.8	51.3	54.9	55	34.6	11.9	15.8	44.4	41.9	14.4	24.1	10.8
Molmo-7B2026.06		33.9	49.1	48.6	49	54.9	55.1	48.1	55.1	54.9	33.3	24.7	26.9	45.1	44.1	22.8	21.4	15