Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Text-to-Image on GenExam Math (Overall)
Loading...
68.2
Strict Accuracy
Draw2Think
-2.728
15.686
34.1
52.514
May 20, 2026
Strict Accuracy
Relaxed Accuracy
Updated 13d ago
Evaluation Results
Method
Method
Links
Strict Accuracy
Relaxed Accuracy
Draw2Think
Model Category=Closed-...
2026.05
68.2
90.5
Nano Banana 2
Model Category=Closed-...
2026.05
56.3
87.8
Nano Banana Pro
Model Category=Closed-...
2026.05
55.6
86.3
GPT-Image-2
Model Category=Closed-...
2026.05
50.3
85.2
Seedream 5.0
Model Category=Closed-...
2026.05
47
82.9
GPT-Image-1.5
Model Category=Closed-...
2026.05
26.5
65.8
Faire
Model Category=Open-so...
2026.05
9.3
52.3
FLUX.2 max
Model Category=Closed-...
2026.05
6.6
49.1
FLUX.2 dev
Model Category=Open-so...
2026.05
2.6
31.6
Qwen-Image-2512
Model Category=Open-so...
2026.05
0
27.9
HunyuanImage-3.0
Model Category=Open-so...
2026.05
0
17
BLIP3o-NEXT-GRPO
Model Category=Open-so...
2026.05
0
15.5
BAGEL
Model Category=Open-so...
2026.05
0
14.7
Janus-Pro
Model Category=Open-so...
2026.05
0
13.7
Feedback
Search any
task
Search any
task