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

Multi-image Spatial Reasoning on MMSI-Bench

38Overall Accuracy

Gemini-2.5-Pro

19.38424.21729.0533.883Jan 12, 2026Jan 16, 2026Jan 20, 2026Jan 24, 2026Jan 28, 2026Feb 1, 2026Feb 5, 2026
Updated 4d ago

Evaluation Results

MethodLinks
2026.01
38----
2026.01
33.7----
2026.01
32.2----
2026.01
32----
2026.01
31----
2026.01
30.3----
2026.01
30.2----
2026.01
30.2----
2026.02
28.829.929.227.326.8
2026.01
28.6----
2026.01
28----
2026.01
26.8----
2026.01
26.1----
2026.02
25.925.921.53025.8
2026.01
20.1----