Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Document Understanding on AI2D (test)
Loading...
88.9
Accuracy
Qwen3-VL 32B
64.1064
70.5432
76.98
83.4168
Jan 29, 2026
Feb 4, 2026
Feb 11, 2026
Feb 18, 2026
Feb 25, 2026
Mar 4, 2026
Mar 11, 2026
Accuracy
Updated 1mo ago
Evaluation Results
Method
Method
Links
Accuracy
Qwen3-VL 32B
Model Source Category=...
2026.01
88.9
Gemini-2.5 Flash
Model Source Category=...
2026.01
88.7
GPT5 mini
Model Source Category=...
2026.01
88.2
MMFineReason-8B
Model Source Category=...
2026.01
87.9
Qwen3-VL 30B-A3B
Model Source Category=...
2026.01
86.9
MMFineReason-4B
Model Source Category=...
2026.01
86.5
HoneyBee 8B
Model Source Category=...
2026.01
86
OMR 7B
Model Source Category=...
2026.01
85
Qwen3-VL 8B
Model Source Category=...
2026.01
84.9
MMR1 8B
Model Source Category=...
2026.01
83.4
MMFineReason-2B
Model Source Category=...
2026.01
82.5
MRoPE
Positional Encoding Me...
2026.03
67.39
MRoPE-I
Positional Encoding Me...
2026.03
67.39
Vanilla RoPE
Positional Encoding Me...
2026.03
66.48
MRoPE-I
Positional Encoding Me...
2026.03
66.26
MRoPE
Positional Encoding Me...
2026.03
65.32
Vanilla RoPE
Positional Encoding Me...
2026.03
65.06
Feedback
Search any
task
Search any
task