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

Spatial Reasoning on MMSI-Bench

97.2Accuracy

Human

Updated 4d ago

Evaluation Results

Method	Links
Human 2026.02		97.2
GEMINI-3-PRO-PREVIEW 2026.02		45.2
GPT-5 2026.02		41.8
SEED-1.6 2026.02		38.3
GEMINI-2.5-PRO 2026.02		38
VST-7B-SFT 2026.02		32
GeoThinker QWEN2.5VL-7B 2026.02		31.7
BAGEL-7B-MOT 2026.02		31
GeoThinker QWEN3VL-8B 2026.02		30.9
VST-3B-SFT 2026.02		30.2
VG-LLM-8B* 2026.02		30
QWEN3-VL-8B-INSTRUCT 2026.02		28.8
QWEN2.5-VL-3B-INSTRUCT 2026.02		28.6
VG-LLM-8B 2026.02		28.4
INTERNVL3-8B 2026.02		28
VG-LLM-4B 2026.02		28
VLM-3R-7B 2026.02		27.9
SPATIALLADDER-3B 2026.02		27.4
QWEN2.5-VL-7B-INSTRUCT 2026.02		26.8
INTERNVL3-2B 2026.02		26.5
CAMBRIAN-S-7B 2026.02		25.8
CAMBRIAN-S-3B 2026.02		25.2
Random Choice 2026.02		25
QWEN3-VL-2B-INSTRUCT 2026.02		11.9