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

Diagram Understanding on AI2D (test)

94.7Accuracy

Claude-3.5-Sonnet

Updated 1d ago

Evaluation Results

Method	Links
Claude-3.5-Sonnet 2024.12		94.7
Gemini-1.5-Pro 2024.12		94.4
GPT-4o 2024.12		94.2
Molmo2-8B 2026.05		94.1
Molmo2-4B 2026.05		93.8
InternVL3.5-8B 2026.05		92.5
Qwen3-VL-8B 2026.05		92.3
PerceptionLM-3B 2026.05		92.2
InternVL3.5-4B 2026.05		92
Qwen3-VL-4B 2026.05		91.8
PerceptionLM-8B 2026.05		91.8
Zamba2-VL-7B 2026.05		90.6
Gemini-2.5-Pro 2026.02		89.5
Bagel 2026.02		89.2
Claude-3 Sonnet 2024.04		88.7
InternVL3.5-2B 2026.05		88.6
InternVL2-76B 2024.12		88.4
Grok-1.5V 2024.04		88.3
Claude-3 Opus 2024.04		88.1
Qwen-VL-Max-0809 2024.10		88.1
Qwen2-VL-72B 2024.12		88.1
InternVL2-Llama3-76B 2024.10		87.6
Qwen2-VL-72B-Thinking 2026.03		87.2
Qwen2-VL-72B-Thinking 2026.03		86.9
Claude-3 Haiku 2024.04		86.7
Qwen3-VL-2B 2026.05		86.2
Manzano 2026.02		86
Zamba2-VL-2.7B 2026.05		85.8
PerceptionLM-1B 2026.05		85.7
LLaVA-OV-72B 2024.12		85.6
LLaVA-OV-72B 2024.12		85.1
VQ-RAE 2026.02		84.8
Phi-4-reasoning-vision-15B 2026.03		84.8
GPT-4o-20240806 2024.10		84.7
GPT-4o 2026.02		84.6
MAmmoTH-VL-8B 2024.12		84
Qwen2.5VL 2026.02		83.9
Qwen2-VL-7B-Thinking 2026.03		83.9
InternVL-2-8B 2024.12		83.8
SAIL-VL1.5 2025.12		83.7
Qwen2-VL-7B-Thinking 2026.03		83.5
Qwen2.5-VL-7B + Cont. Reward 2025.11		83.41
MAmmoTH-VL-8B 2024.12		83.4
HyperVL ViTL 2025.12		83.1
Qwen2-VL-7B-Ins 2024.12		83
SAIL-VL2 2025.12		83
Ming-Univision 2026.02		82.8
Vision-Zero 2025.11		82.64
Qwen2.5-VL-7B (Baseline) 2025.11		82.61
InternVL3.5 2026.02		82.6
Kelix 2026.02		82.4
Qwen2.5-VL-7B + Discrete Reward 2025.11		82.18
MiniCPM-V-2.6-7B 2024.12		82.1
HyperVL 2025.12		81.8
Circle-RoPE 2025.05		81.8
LLaVA-OV-7B 2024.12		81.6
Qwen2.5VL 2026.02		81.6
Qwen2.5-VL-3B-Instruct 2025.12		81.51
INXComp-2.5-7B 2024.12		81.5
Zamba2-VL-1.2B 2026.05		81.5
LLaVA-OV-7B 2024.12		81.4
LLaVA-OV 2026.02		81.4
InternVL2.5 2025.05		81.38
MiniCPM V-2.6 2025.05		81.28
Kimi-VL-A3B-Thinking 2026.03		81.2
InternVL3.5-1B 2026.05		81.2
Molmo-7B-D 2024.12		81
InternVL 1.5 2024.04		80.7
gemma-2-9b-it 2026.03		80.4
Gemini Pro 1.5 2024.04		80.3
Claude3.5-Sonnet 2024.10		80.2
AndesVL 2025.12		80.1
Ovis2 2025.12		79.7
Phi-4-reasoning-vision-15B 2026.03		79.7
Gemini Ultra 1.0 2024.04		79.5
Qwen-VL-Max 2024.04		79.3
TUNA 2026.02		79.3
Qwen2.5-VL (SFT) 2025.05		79.22
Step-1V 2024.04		79.2
Gemini-1.5-Pro 2024.10		79.1
InternVL 1.2 2024.04		79
InternVL3.5 2025.12		78.8
InternVL3 2025.12		78.7
MobileNet-Qwen 2025.12		78.63
Show-O2 2026.02		78.6
Phi-3-Vision 2024.10		78.4
GPT-4V 2024.04		78.2
Qwen2.5-VL 2025.05		78.14
SAIL-VL 2025.05		77.72
Phi-3.5 vision 2025.05		77.59
Llama-3.2-11B-Vision-Ins 2024.12		77.3
InternViT-Qwen 2025.12		77.17
Qwen3-VL 2025.12		76.9
X-Omni 2026.02		76.8
Qwen-VL-Plus 2024.04		75.9
TokenFlow-XL 2026.02		75.8
LLaVA-NeXT 2024.04		74.9
Qwen2-VL-2B 2024.10		74.7
Qwen2-VL 2025.12		74.7
Qwen2-VL-2B 2025.12		74.64

Showing 100 of 154 rows