Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Multimodal Image Understanding on DocVQA
Loading...
96.1
Score
Qwen3-VL
22.4368
41.5609
60.685
79.8091
Oct 27, 2025
Dec 4, 2025
Jan 12, 2026
Feb 20, 2026
Mar 30, 2026
May 8, 2026
Jun 16, 2026
Score
Updated 1mo ago
Evaluation Results
Method
Method
Links
Score
Qwen3-VL
Category=Und. Only Models
2026.06
96.1
UniAR
Category=Unified Models
2026.06
91.4
X-Omni
Category=Unified Models
2026.06
88.6
LLaVA-OV
Category=Und. Only Models
2026.06
87.5
Emu3
Category=Unified Models
2026.06
76.3
LLaVA-1.6
Category=Und. Only Models
2026.06
74.4
MRoPE-I
Backbone Model=Qwen3-V...
2025.10
46.89
MHRoPE
Backbone Model=Qwen3-V...
2025.10
46.77
Vanilla RoPE
Backbone Model=Qwen3-V...
2025.10
46.59
CircleRoPE
Backbone Model=Qwen3-V...
2025.10
44.6
MRoPE
Backbone Model=Qwen3-V...
2025.10
43.13
HoPE
Backbone Model=Qwen3-V...
2025.10
27.25
VideoRoPE
Backbone Model=Qwen3-V...
2025.10
25.27
Feedback
Search any
task
Search any
task
