Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Document/Chart Understanding on DocVQA
Loading...
94.9
Score
Qwen2.5-VL
54.548
65.024
75.5
85.976
Dec 22, 2025
Jan 20, 2026
Feb 19, 2026
Mar 20, 2026
Apr 19, 2026
May 18, 2026
Jun 17, 2026
Score
Updated 1mo ago
Evaluation Results
Method
Method
Links
Score
Qwen2.5-VL
Model Size=7B
2026.06
94.9
Dream-VL
Model Size=7B
2026.06
94.4
Qwen2.5-VL 3B (reported)
Res.=Native
2025.12
93.6
InternVL3
Model Size=8B
2026.06
92.7
Qwen2.5-VL 3B (reproduced)
Res.=≤ 896²
2025.12
91.5
PerceptionDLM-Base
Model Size=8B
2026.06
89.9
CASA⊕ Qwen2.5-VL
Res.=≤ 896²
2025.12
88.9
SDAR-VL
Model Size=8B
2026.06
88.3
LLaDA-V
Model Size=8B
2026.06
83.9
LaViDa
Model Size=8B
2026.06
56.1
Feedback
Search any
task
Search any
task