List your company

Static Spatial Reasoning on STI-Bench Static

37.1Avg. ScoreGemini-2.5-Pro
21.60425.62729.6533.673Jun 10, 2026

Evaluation results

MethodLinks
Gemini-2.5-ProType=Proprietary Model (API)2026.0637.134.253.432.3
Claude-3.7-SonnetType=Proprietary Model (API)2026.063731.84936.3
Gemini-2.0-FlashType=Proprietary Model (API)2026.0636.933.75033.7
ReReBackbone=INTERNVL2.5-8B2026.0634.833.249.329.7
ReReBackbone=QWEN3-VL-4B2026.0634.433.648.628.7
INTERNVL2.5-8BFramework=Baseline2026.0632.130.145.927.8
GPT-4oType=Proprietary Model (API)2026.063124.949.628.1
ReReBackbone=QWEN3-VL-8B2026.0630.929.144.526.2
ReReBackbone=INTERNVL2.5-4B2026.0630.622.243.232.5
INTERNVL2.5-4BFramework=Baseline2026.0630.525.943.129
ReReBackbone=QWEN3-VL-2B2026.0630.224.65026.2
QWEN3-VL-4BFramework=Baseline2026.0629.729.841.824.3
QWEN3-VL-8BFramework=Baseline2026.0627.929.443.819.2
ReReBackbone=INTERNVL3-8B2026.0627.8283723.3
ReReBackbone=INTERNVL3-2B2026.0626.722.543.822.7
INTERNVL3-8BFramework=Baseline2026.0624.520.144.519.2
INTERNVL3-2BFramework=Baseline2026.0622.620.128.822.1
QWEN3-VL-2BFramework=Baseline2026.0622.21831.521.8