Spatial Reasoning on VSI-Bench 59 (test) benchmark leaderboard · SOTA2 Research

Spatial Reasoning on VSI-Bench 59 (test)

72.5Object Count ScoreMotion-MLLM-4B

Evaluation Results

Method	Links
Motion-MLLM-4B2026.03		72.5	49.2	69.5	66.9	57.8	56.7	40.5	59.6	60.3
InternVL3-78B2026.03		71.2	53.7	44.4	39.5	55.9	39.5	28.9	54.5	48.5
VG LLM-8B2026.03		67.9	37.7	58.6	62	46.6	40.7	32.4	59.2	50.7
VG LLM-4B2026.03		66	37.8	55.2	59.2	44.6	45.6	33.5	36.4	47.3
Spatial-MLLM-4B2026.03		65.3	34.8	63.1	45.1	41.3	46.2	33.5	46.3	48.4
Spacer2026.03		57.8	28.2	59.9	47.1	40.1	45.4	33.5	52.1	45.5
Gemini-1.5 Pro2026.03		56.2	30.9	64.1	43.6	51.3	46.3	36	34.6	45.4
LLaVA-Video-72B2026.03		48.9	22.8	57.4	35.3	42.4	36.7	35	48.6	40.9
GPT-4o2026.03		46.2	5.3	43.8	38.2	37	41.3	31.5	28.5	34
LLaVA-OneVision-72B2026.03		43.5	23.9	57.6	37.5	42.5	39.9	32.5	44.6	40.2
Qwen2.5-VL-7B2026.03		40.9	14.8	43.4	10.7	38.6	38.5	33	29.8	33
Qwen2.5-VL-72B2026.03		25.1	29.3	54.5	38.8	38.2	37	34	28.9	37
Qwen2.5-VL-3B2026.03		24.3	24.7	31.7	22.6	38.3	41.6	26.3	21.2	30.6
VILA-1.5-40B2026.03		22.4	24.8	48.7	22.7	40.5	25.7	31.5	32.9	31.2