Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Multimodal Understanding on AI2D w/o M
Loading...
81.6
Accuracy
Qwen2.5VL-3B
45.408
54.804
64.2
73.596
Dec 9, 2025
Accuracy
Updated 1mo ago
Evaluation Results
Method
Method
Links
Accuracy
Qwen2.5VL-3B
Complexity=O(n)
2025.12
81.6
InternVL2.5-4B
Complexity=O(n)
2025.12
81.4
Phi-3.5-Vision-4B
Complexity=O(n)
2025.12
77.8
InfiniteVL-4B
Complexity=O(1)
2025.12
77.2
SmolVLM2-2B
Complexity=O(n)
2025.12
74.9
PaliGemma2-3B
Complexity=O(n)
2025.12
72.2
TinyLLaVA-3B
Complexity=O(n)
2025.12
61.8
MaTVLM-3B
Complexity=O(n)
2025.12
58.9
Cobra-3B
Complexity=O(1)
2025.12
46.8
Feedback
Search any
task
Search any
task