Share your thoughts, 1 month free Claude Pro on usSee more

Document Visual Question Answering on DocVQA v1.0 (test)

96.5ANLS

Qwen2-VL-72B

Updated 1d ago

Evaluation Results

Method
Qwen2-VL-72B 2024.12	96.5	-
InternVL2.5-38B 2024.12	95.3	-
Claude-3.5-Sonnet 2024.12	95.2	-
InternVL2.5-78B 2024.12	95.1	-
Qwen2-VL-7B 2024.12	94.5	-
InternVL2-Llama3-76B 2024.12	94.1	-
InternVL2.5-26B 2024.12	94	-
InternVL2-40B 2024.12	93.9	-
Molmo-72B 2024.12	93.5	-
Gemini-1.5-Pro 2024.12	93.1	-
InternVL2.5-8B 2024.12	93	-
InternVL2-26B 2024.12	92.9	-
GPT-4o-20240513 2024.12	92.8	-
NVLM-D-72B 2024.12	92.6	-
InternVL-1.5-Plus 2024.11	92.3	-
Molmo-7B-D 2024.12	92.2	-
InternVL2.5-4B 2024.12	91.6	-
InternVL2-8B 2024.12	91.6	-
LLaVA-OneVision-72B 2024.12	91.3	-
InternVL-Chat-V1.5 2024.12	90.9	-
MiniCPM-V2.6 2024.12	90.8	-
Qwen2-VL-2B 2024.12	90.1	-
Claude-3-Opus 2024.12	89.3	-
InternVL2-4B 2024.12	89.2	-
InternVL2.5-2B 2024.12	88.7	-
GPT-4V 2024.12	88.4	-
InternVL2-2B 2024.12	86.9	-
Aquila-VL-2B 2024.12	85	-
InternVL2.5-1B 2024.12	84.8	-
LayoutLMv3 2024.04	83.37	-
InternVL2-1B 2024.12	81.7	-
Cambrian-34B 2024.12	75.5	-
LayoutLLM-7B 2024.04	74.27	-
LayoutLLM-7B 2024.04	74.25	-
HVFA 2024.11	72.7	-
LLaVA-OneVision-0.5B 2024.12	70	-
Phi-3.5-Vision-4B 2024.12	69.3	-
Vicuna-1.5-7B 2024.04	66.99	-
Qwen-VL-7B 2024.04	65.1	-
Llama2-7B-chat 2024.04	64.99	-
mPLUG-DocOWL-7B 2024.04	62.2	-
Vicuna-7B 2024.04	61.39	-
Llama2-7B 2024.04	61.34	-
Vicuna-1.5-7B 2024.04	56.81	-
Llama2-7B-chat 2024.04	56.55	-
Llama2-7B 2024.04	37.32	-
Vicuna-7B 2024.04	37.21	-
LLaVA-1.5-7B 2024.04	13.34	-
LLaVAR-7B 2024.04	11.6	-
LLaVA-1.5 7B 2026.02	-	22.31
LLaVA-1.5 7B + VIG training 2026.02	-	22.51
LLaVA-1.5 13B 2026.02	-	24.08
LLaVA-1.5 13B + VIG training 2026.02	-	25.27
ShareGPT4V 7B 2026.02	-	26.15
ShareGPT4V 7B + VIG training 2026.02	-	28.23
GPT-5.4 2026.06	-	96.85
Claude-Sonnet-4-6 2026.06	-	98.18
Gemini-3.1-Pro-Preview 2026.06	-	96.95
Qwen2.5-VL-7B 2026.06	-	93.43
Qwen3-VL-8B 2026.06	-	89.8
Qwen3-VL-32B 2026.06	-	94.7
InternVL3.5-14B 2026.06	-	87.6
InternVL3.5-38B 2026.06	-	91.8
VTool-R1-7B 2026.06	-	93.1
R1-Onevision-7B 2026.06	-	83.2
PyVision-Image-7B-RL 2026.06	-	69.95
DeepEyesV2-7B 2026.06	-	89.9
PixelReasoner-RL-v1-7B 2026.06	-	92.3
AdaReasoner-7B 2026.06	-	89.4
TikArt-8B 2026.06	-	96.7
ZwZ-8B 2026.06	-	96.8
MetaForge 2026.06	-	97.9
GPT-5.4 2026.06	-	96.55
Claude-Sonnet-4-6 2026.06	-	98.08
Gemini-3.1-Pro-Preview 2026.06	-	96.13
Qwen2.5-VL-7B 2026.06	-	93.63
Qwen3-VL-8B 2026.06	-	90.2
Qwen3-VL-32B 2026.06	-	95.15
InternVL3.5-14B 2026.06	-	88.2
InternVL3.5-38B 2026.06	-	92.1
VTool-R1-7B 2026.06	-	93.25
R1-Onevision-7B 2026.06	-	83.9
PyVision-Image-7B-RL 2026.06	-	77
DeepEyesV2-7B 2026.06	-	93.05
PixelReasoner-RL-v1-7B 2026.06	-	91.85
AdaReasoner-7B 2026.06	-	92.3
TikArt-8B 2026.06	-	94.7
ZwZ-8B 2026.06	-	95.6
MetaForge 2026.06	-	97.22