Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
Edit Inspection Questions on Imagen3 edits
Loading...
58.8
Accuracy
LLaVA
48.816
51.408
54
56.592
Jun 11, 2025
Accuracy
Contextual Consistency
Technical Precision
Artifacts Score
Diff Caption Acc
Updated 4d ago
Evaluation Results
Method
Method
Links
Accuracy
Contextual Consistency
Technical Precision
Artifacts Score
Diff Caption Acc
LLaVA
training=supervised
2025.06
58.8
-
-
48.8
53.9
GPT-4
2025.06
55.3
47.7
51.1
47.1
54.2
LLaVA
2025.06
54.4
-
-
43.6
50
Qwen2.5 VL
2025.06
52.9
42.1
47.5
50
51.8
InternVL3
2025.06
51.5
48.9
54.5
48.6
50.8
GPT-4o
2025.06
49.9
49.5
50.9
56.5
58.2
GPT-4 Turbo
2025.06
49.2
55.6
49.1
49.5
55.9
Feedback
Search any
task
Search any
task