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

Visual Question Answering on VQAv2

97.4Accuracy

Oracle

Updated 7d ago

Evaluation Results

Method	Links
Oracle 2026.05		97.4
SEA-PRIME 2024.08		83.1
MM1 2025.01		82.8
PIIP-LLaVA 2025.01		82.5
LLAVA-HR 2025.01		82.3
PIIP-LLaVA 2025.01		82.3
CogVLM 2024.08		82.3
MM1-7B-Chat 2024.06		82.3
mPLUG-Owl3 2024.08		82.1
MG-LLAVA 2024.06		82
LLaVA-HR 2025.01		81.9
SEA-PRIME 2024.08		81.9
EVLM-Chat 2024.08		81.9
PIIP-LLaVA 2025.01		81.8
LLaVA-NeXT 2025.01		81.8
LLAVA-Next 2024.08		81.8
LLaVA-NeXT 2024.06		81.8
LLaVA-UHD 2024.06		81.7
PIIP-LLaVA 2025.01		81.5
SEA-PRIME 2024.08		81.4
VisionLLM v2-chat 2024.06		81.4
Flamingo 2024.08		81.3
Mipha-3B 2024.12		81.3
MG-LLAVA 2024.06		81.2
PIIP-LLaVA 2025.01		81.1
SEA-PRIME 2024.08		81
Dragonfly (Ours) 2024.06		81
S2-Wrapper 2024.08		80.9
VILA-1.5 2024.08		80.9
VILA-13B 2023.12		80.8
Idefics2 2024.08		80.8
DualPD 2026.01		80.78
SEA-PRIME 2024.08		80.7
MG-LLAVA 2024.06		80.7
SPHINX-2k 2024.06		80.7
ShareGPT4V 2024.12		80.6
ShareGPT4V 2024.08		80.6
VILA-13B + ShareGPT4V 2023.12		80.6
ShareGPT4V-7B 2024.06		80.6
Cloud-Only 2026.05		80.6
Olympus 2024.12		80.5
Slime 2025.01		80.3
Monkey 2024.06		80.3
MG-LLaVA 2025.01		80.2
SPHINX-1k 2024.06		80.2
MG-LLAVA 2024.06		80.2
HiMAP 2025.03		80.2
LLaVA-1.5-13B + MMFuser 2024.10		80.1
LLaVA-1.5 2025.01		80
LLaVA-1.5 2024.12		80
Show-o 2024.12		80
LLaMA-VID 2024.08		80
LLaVA-1.5* 2024.08		80
AlignGPT 2024.08		80
LLaVA-1.5 2023.12		80
LLaVA-1.5 2024.06		80
LLaVA-1.5-13B 2024.10		80
Full-Finetune 2026.05		80
TinyLLaVA 2024.08		79.9
VILA-7B 2023.12		79.9
VILA 2024.06		79.9
MoE-LLaVA-3.6B 2024.12		79.9
TinyLLaVA 2024.12		79.9
FastV 2025.03		79.9
Visual Prompt 2024.08		79.8
Bunny-3B 2024.12		79.8
Baseline 2025.03		79.8
S2-Wrapper* 2024.08		79.7
HiMAP 2025.03		79.6
Qwen-VL 2024.08		79.5
Imp-v1 2024.12		79.5
mPLUG-Owl2 2025.01		79.4
VILA-U 2024.12		79.4
mPLUG-Owl2 2024.08		79.4
mPLUG-Owl2 2024.06		79.4
mPLUG-Owl2 2024.12		79.4
Baseline 2025.03		79.4
LLaMA-VID 2024.08		79.3
InternVL-7B 2024.06		79.3
Qwen-2.5-VL-7B 2026.01		79.3
AlignGPT 2024.08		79.1
LLaVA-1.5-7B + MMFuser 2024.10		79.1
INAR-VL 2026.05		79.1
FastV 2025.03		79
MAGIC 2026.05		78.9
DoLA 2026.01		78.85
LLaVA-1.5* 2024.08		78.8
Qwen-VL 2023.12		78.8
Qwen-VL 2024.06		78.8
Qwen-VL 2024.10		78.8
Qwen-VL 2024.11		78.8
HiMAP 2025.03		78.8
Text-Only 2026.05		78.8
HiMAP 2025.03		78.6
DoLA 2026.01		78.57
LLaVA-1.5 2025.01		78.5
LLaVA-1.5 2023.12		78.5
LLaVA-1.5 2024.06		78.5
LLAVA-1.5 2024.08		78.5
LLaVA-1.5 2024.06		78.5

Showing 100 of 196 rows