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

Visual Question Answering on XLRS-Bench L-3 Capability (test)

41.7OC

Gemini 2.0 Flash

Updated 3d ago

Evaluation Results

Method	Links
Gemini 2.0 Flash 2026.04		41.7	45	38	73.5	34.6	27.6	61.7	32	73	82	43	30	51	39.5
InternVL3-8B 2026.04		40	39	10	71.5	44.5	30.8	65	25.2	77	82	36	21.7	50	40.3
InternVL2.5-8B 2026.04		38.3	37	10	77	33.4	35.5	65	21.6	73	83	34	50	43	38.5
Qwen2.5-VL-7B 2026.04		33.3	40	31	77	40.6	40.5	66.7	36.2	68	72	27	38.3	45	43.8
LLaVA-OneVision-72B 2026.04		33.3	38	15	72.5	36.3	36.3	66.7	35.6	74	83	28	36.7	43	41.1
Claude 3.7 Sonnet 2026.04		27.6	22.7	17.4	68.4	30.5	29.9	63.6	27.6	64.8	78.4	34.5	27.8	32.6	35.1
LLaVA-Next 2026.04		26.7	40	5	67	28.8	32.8	66.7	30	69	78	27	35	36	36
Qwen2-VL-7B 2026.04		26.7	40	11	73	35.9	34.6	61.7	31.8	70	81	35	46.7	48	40.1
GeoLLaVA-8K 2026.04		26.7	38	49	69	41.6	31.6	65	35	67	78	66	50	52	43.3
LLaVA-OneVision-7B 2026.04		25	38	2	69.5	35.9	35.3	65	25.2	76	83	24	43.3	36	38.1
GPT-4o 2026.04		25	32	15	66	9.5	11.3	11.7	24.6	73	73	35	20	25	23
InternVL3-78B 2026.04		23.3	49	33	74	42.5	37.4	66.7	30	76	81	40	45	42	43.5
GPT-4o-mini 2026.04		23.3	25	19	59.5	40.9	31	65	23.6	71	71	29	6.7	30	36.2
InternLM-XComposer-2.5 2026.04		21.7	42	7	68	31.8	27.8	6.7	26	72	81	41	36.7	47	34.8
UHR-BAT 2026.04		21.7	33	50	55.5	43.5	33.8	65	44.8	62	71	54	46.7	51	44
EarthDial 2026.04		18.3	42	1	36	31.3	31	65	24.8	62	71	43	48.3	50	33.8
GeoChat 2026.04		16.7	29	2	23	21.1	16.8	35	24.2	33	43	10	24	21	21.2
VHM 2026.04		16.7	30	2	26	21.4	16.8	35	25.6	42	53	46	28.3	21	23.6