Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
Multimodal Autoformalization on SYNTHETIC GEOMETRY Plane Geometry
Loading...
70
Compile Rate
Gemini-3-Pro
-2.8
16.1
35
53.9
Jan 6, 2026
Compile Rate
Semantic Correctness
Human Check Score
Updated 3d ago
Evaluation Results
Method
Method
Links
Compile Rate
Semantic Correctness
Human Check Score
Gemini-3-Pro
Modality=Image
2026.01
70
70
40
GPT-5
Modality=Image
2026.01
40
40
20
Gemini-3-Pro
Modality=Text
2026.01
40
30
-
GPT-5
Modality=Text
2026.01
30
30
-
Gemini-2.5-Pro
Modality=Image
2026.01
30
30
20
Gemini-2.5-Pro
Modality=Text
2026.01
30
30
-
Qwen3-VL-235B
Modality=Text
2026.01
10
10
-
Qwen2.5-VL-72B
Modality=Image
2026.01
10
0
0
Qwen3-VL-235B
Modality=Image
2026.01
0
0
0
Qwen2.5-VL-72B
Modality=Text
2026.01
0
0
-
Feedback
Search any
task
Search any
task