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

Spatial Reasoning (Multi-Image) on ERQA

51.02Accuracy

Gemini2.5-Pro

Updated 9d ago

Evaluation Results

Method	Links
Gemini2.5-Pro 2025.11		51.02
MiMo-Embodied 2025.11		46.75
Claude-Sonnet-4 2025.11		45.75
ViLAVT-7B 2026.02		42.2
MiMo-VL 2025.11		41.5
VILASR-7B 2026.02		41
InternVL3.5 2025.11		41
Qwen-VL-Max 2025.11		40.91
SpaceR-7B 2026.02		40.3
Qwen2.5-VL-7B 2026.02		39.8
Qwen2.5-VL-7B 2026.02		39.3
Cosmos-Reason1 2025.11		39.09
Qwen2.5-VL 2025.11		38.8
Spatial-MLLM-4B 2026.02		38.3
VeBrain 2025.11		37.29
RoboBrain-1.0 2025.11		36.52
InternVL3-8B 2026.02		35.3
GPT-4o 2025.11		32.48
LLaVA-OneVision-7B 2026.02		30.6
RoboBrain-2.0 2025.11		30.31
Magma 2025.11		25.73
Multi-SpatialMLLM 2025.05		21.6
InternVL-8B 2025.05		13.5