Choice-based Visual Question Answering on EgoPoint-Bench simulation 1.0 (test) benchmark leaderboard · SOTA2 Research

Choice-based Visual Question Answering on EgoPoint-Bench simulation 1.0 (test)

91.86BP ScoreHuman

Evaluation Results

Method	Links
Human2026.04		91.86	97.14	100	93.33	100	95.8
Qwen3-VL-8B2026.04		81.31	80.92	80.56	84.24	82.91	81.36
Qwen3-VL-32B2026.04		80.75	82.5	83.39	83.03	82.84	82.2
LLaVA-NeXT-7B2026.04		80.39	80.86	79.56	72.42	86.13	80.93
GLM-4.6V-Flash2026.04		77.16	73.28	82.01	80	64.21	74.86
LLaVA-1.5-7B2026.04		76.41	72.06	60.63	66.06	86.44	73.18
InternVL3.5-14B2026.04		75.99	76	83.01	76.36	86.51	78.59
InternVL3.5-8B2026.04		74.6	77.81	82.76	78.79	86.21	78.86
InternVL3.5-2B2026.04		71.4	75.36	76.61	78.79	81.99	75.43
GPT-5 mini2026.04		59.96	58.22	67.65	68.79	36.09	57.66
Qwen3-VL-8B2026.04		57.55	54	70.34	77.58	52.11	58.29
Qwen3-VL-32B2026.04		56.52	53.75	65.64	79.39	60.23	58.28
GLM-4.6V-Flash2026.04		56.16	50.81	66.14	61.52	36.17	53.29
Gemini 3 Flash2026.04		54.39	53.33	66.58	73.64	58.39	57.21
GPT-5.2 Instant2026.04		54.14	49.81	66.14	75.45	50.88	54.8
InternVL3.5-8B2026.04		52.86	52.5	63.51	66.36	35.63	52.62
Gemini 3 Pro2026.04		52.47	51.39	70.47	74.85	57.16	56.44
InternVL3.5-2B2026.04		51.97	55.14	61.5	66.97	26.05	51.74
LLaVA-1.5-7B2026.04		50.83	46.89	54.86	50.91	41.92	48.82
LLaVA-NeXT-7B2026.04		47.42	45.42	55.92	53.33	46.59	48.17
InternVL3.5-14B2026.04		46.79	51.14	62.07	71.52	33.56	49.99
Random2026.04		27.95	26.83	38.89	43.24	52.17	31.14