Spatial Reasoning
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
99.9Binary Robust Accuracy
9
Jun 2, 2026
100Binary Robust Accuracy
9
Jun 2, 2026
72RDB (Back)
9
May 12, 2026
67Accuracy
9
Mar 12, 2026
72.4Accuracy
9
Mar 12, 2026
82.11Confidence
9
Mar 5, 2026
95.7Open-ended Score
9
Feb 26, 2026
89.7Open-ended Score
9
Feb 26, 2026
99.51Ctrl-A Score
8
May 19, 2026
44.4Score
8
Apr 21, 2026
66Relative Position Accuracy
8
Mar 16, 2026
47.46Overall Score
8
Mar 10, 2026
71.65Accuracy
8
Mar 10, 2026
53.96Accuracy
8
Mar 10, 2026
55.05Average Accuracy
8
Mar 4, 2026
43.8Accuracy
8
Feb 26, 2026
52.9Accuracy
8
Feb 26, 2026
70.2Accuracy
8
Feb 26, 2026
60.7Accuracy
8
Feb 26, 2026
92.7Accuracy
8
Feb 26, 2026
62.7Average Score
8
Feb 26, 2026
77.5HR-4K Average Score
7
Apr 23, 2026
92.5Average Score
7
Apr 23, 2026
98Accuracy
7
Mar 17, 2026
62.26Accuracy
7
Mar 17, 2026