Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
Human Preference Evaluation on PhysicsIQ 1.0 (test)
Loading...
54.9
Physics Plausibility Win Rate
MAGI-1
44.708
47.354
50
52.646
Jan 15, 2026
Physics Plausibility Win Rate
Physics Plausibility Accuracy
Visual Quality Win Rate
Visual Quality Accuracy
Prompt Alignment Win Rate
Prompt Alignment Accuracy
Overall Win Rate
Overall Accuracy
Updated 4d ago
Evaluation Results
Method
Method
Links
Physics Plausibility Win Rate
Physics Plausibility Accuracy
Visual Quality Win Rate
Visual Quality Accuracy
Prompt Alignment Win Rate
Prompt Alignment Accuracy
Overall Win Rate
Overall Accuracy
MAGI-1
Sampling Strategy=WMRe...
2026.01
54.9
58.7
52.9
57.1
55.7
64.9
54.5
60
vLDM
Sampling Strategy=WMRe...
2026.01
53.1
54.8
54.7
58.8
51.3
53.2
53
55.7
vLDM
Sampling Strategy=vanilla
2026.01
46.9
45.2
45.3
41.2
48.7
46.8
47
44.3
MAGI-1
Sampling Strategy=vanilla
2026.01
45.1
41.3
47.1
42.9
44.3
35.1
45.5
40
Feedback
Search any
task
Search any
task