Our new X account is live! Follow @wizwand_team for updates

Spatial Reasoning on Escher-Bench 1.0 (test)

56.17Object Permanence & Occlusion Tracking

Gemini-2.5-pro

Updated 3d ago

Evaluation Results

Method
Gemini-2.5-pro 2026.01	56.17	57.29	69.53	54.2	73.11	54.39	57.26	58.37	53.64	64.53	59.6	51.57	60.74
Gemini-2.5-flash 2026.01	53.13	51.71	65.01	49.22	65.88	48.91	52.19	52.93	49.66	56.02	48.55	45.51	58.42
GPT5 2026.01	52.24	53.28	64.66	50.41	67.99	50.58	53.36	54.28	49.88	60.01	55.43	47.96	56.49
Escher-4B-Instruct 2026.01	49.72	48.59	61.63	47.33	52.61	47.71	49.49	49.73	48.71	51.82	41.87	47	52.41
Qwen3-VL-32B-Thinking 2026.01	49.26	49.8	58.76	49.34	60.19	46.04	49.58	49.7	49.18	57.68	48.55	43.3	53.87
Qwen3-VL-30B-A3B-Instruct 2026.01	48.53	44.46	54.38	43.08	56.87	43.7	45.73	46.35	43.75	47.79	37.86	39.04	53.05
Escher-4B-Thinking 2026.01	48.07	47.55	61.33	47.32	53.08	46.38	48.54	48.86	47.49	49.72	43.65	45.21	52.26
Escher-7B-Instruct 2026.01	47.61	48.88	62.08	48.89	53.55	45.31	48.89	49.17	47.97	51.05	41.2	46.88	51.38
Escher-8B-Instruct 2026.01	46.96	49.8	63.6	48.31	56.87	47.5	49.85	50.88	46.55	52.6	48.11	46.08	53.17
Qwen3-VL-8B-Instruct 2026.01	46.69	43.98	55.89	43.08	58.29	42.21	45.06	45.74	42.86	47.79	38.31	37.82	52.6
ISI-8B-Instruct 2026.01	46.59	43.01	56.5	43.46	50.24	43.64	45.06	46.06	41.86	50.61	26.5	38.92	52.32
Escher-8B-Thinking 2026.01	46.04	48.67	62.84	48.64	60.19	45.22	48.79	49.12	47.71	50.28	44.99	46.08	51.66
Qwen3-VL-4B-Instruct 2026.01	44.57	43.74	48.49	43	52.61	43.08	44.18	44.9	41.86	40.88	35.63	38.74	51.78
qwen3-vl-plus 2026.01	44.38	43.21	55.52	44.61	57.82	42.38	44.75	45.83	41.28	44.86	44.77	38.63	50.94
MiniCPM-V-4 2026.01	43.19	38.67	46.53	37.94	32.23	38.54	39.62	39.81	39.06	35.25	22.27	36.65	46.22
Qwen3-VL-4B-Thinking 2026.01	43.09	43.32	49.85	43.62	54.5	44.63	44.55	46.01	39.85	52.15	35.19	39.04	49.35
LLaVA-OneVision-1.5-4B-Instruct 2026.01	42.82	43.35	48.64	41.52	50.71	43.64	43.73	44.27	41.96	44.64	36.3	38.56	49.74
LLaVA-OneVision-1.5-8B-Instruct 2026.01	41.71	44.87	48.94	42.59	49.29	44.67	44.5	45.22	42.17	42.87	36.75	39.34	51.26
ERNIE-4.5-VL-28B-A3B-Thinking 2026.01	41.34	40.17	51.66	40.28	57.35	41.22	42.58	43.11	41.08	49.72	35.19	37.85	49.01
Qwen3-VL-8B-Thinking 2026.01	41.25	44.26	51.21	41.02	58.29	44.45	44.36	45.7	40.06	51.82	41.43	38.62	48.56
Qwen2.5-VL-7B-Instruct 2026.01	39.32	35.73	45.17	38.02	38.39	33.75	36.84	36.69	37.32	26.08	12.92	32.39	47.59
Qwen2.5-VL-3B-Instruct 2026.01	39.13	38.98	47.13	41.89	44.55	40.15	40.61	41.12	38.96	36.41	11.36	35.46	50.99
Qwen3-VL-2B-Instruct 2026.01	37.75	37.53	44.11	38.35	41.23	38.63	38.65	39.62	35.53	31.71	7.8	34.27	49.23
Spatial-MLLM 2026.01	10.6	15.2	26.13	22.8	11.37	12.81	15.82	15.6	16.55	16.02	1.11	33.28	0
ViCA2-7B 2026.01	9.67	15.24	27.79	23.97	14.69	13.92	16.45	15.96	18.03	17.57	2	34.21	0.03
LLaVA-NeXT-Video-7B 2026.01	6.26	10.28	8.91	16.05	2.84	13.66	11.3	10.63	13.44	1.1	6.46	25.8	0