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

Diagram Understanding on AI2D

94.2Accuracy

GPT-4o

Updated 17d ago

Evaluation Results

Method	Links
GPT-4o 2024.08		94.2
InternVL2.5-8B-GenRecal 2025.06		93
LaVer 2025.12		92.71
InternVL2.5-8B-GenRecal 2025.06		92.2
InternVL2.5-8B-GenRecal 2025.06		92.1
Gemini 2.5 Pro 2026.02		90.9
Bagel-7B-MoT 2025.11		89.5
InternVL2.5-78B 2025.06		89.1
LaVer 2025.12		89.09
VLSI-2B 2024.12		89
VLSI-2B 2024.12		89
SenseNova-SI 2025.11		88.8
Claude-3-Sonnet 2026.02		88.7
InternVL2.5-8B-GenRecal 2025.06		88.5
Qwen2-VL-72B 2024.12		88.3
Grok-1.5V 2026.02		88.3
Claude-3-Opus 2026.02		88.1
Qwen2-VL-72B 2025.06		88.1
Baseline 2025.12		88
InternVL3.5-38B 2025.12		87.8
Qwen2.5-VL-32B 2026.03		87.8
InternVL2-76B 2024.12		87.6
InternVL2-76B 2025.06		87.6
LaVer 2025.12		87.46
VLSI-7B 2024.12		87.3
Baseline 2025.12		86.51
LLaVA-OV-72B 2024.12		86.2
R1-ShareVL-32B 2026.03		86.2
Penguin-VL 2026.03		86.1
Baseline 2025.12		86.02
Qwen3-VL 2026.03		85.7
Qwen2.5-VL-32B 2026.03		85.7
LLaVA-OneVision-72B 2024.08		85.6
LLaVA-OneVision-72B 2025.06		85.6
Qwen2.5-VL-7B 2026.03		85.6
NVLM-72B 2025.06		85.2
InternVL3-8B 2025.11		85.2
VST-7B-SFT 2025.11		84.9
GPT-4o (0806) 2024.12		84.7
GPT-4o-0513 2024.09		84.6
GPT4o 2026.02		84.6
GPT-4o (0513) 2025.06		84.6
Qwen2.5-VL-32B 2026.03		84.6
R1-ShareVL-7B 2026.03		84.5
Qwen2.5-VL-7B 2026.03		84.5
SenseNova-SI 2025.11		84.2
Qwen3-VL + DiG 2025.12		84.1
InternVL-3.5 2026.03		84
Qwen2.5-VL 72B 2025.12		83.9
Qwen2.5-VL-7B 2026.03		83.9
InternVL2 2026.03		83.8
InternVL2-8B 2024.09		83.6
Molmo-72B 2024.12		83.4
Molmo-72B 2025.06		83.4
Perceptio 2026.03		83.4
LaVer 2025.12		83.3
SAYO-Qwen-8B 2026.02		83.06
ThinkLite-7B 2026.03		83
OpenVLThinker-7B 2026.02		82.61
Ovis1.5-LLAMA3-8B 2024.09		82.5
Qwen3-VL 2025.12		82.5
OneVision 2024.09		82.4
GPT-4o 2024.06		82.2
Qwen3-VL + DiG 2025.12		82.2
Qwen3-VL-8B-Thinking 2026.02		82.2
Xuanwu VL-2B 2026.03		82.19
Sa2VA 2026.03		82.1
Semantic-back-7B 2026.02		81.74
Baseline 2025.12		81.61
IXC-2.5 2024.09		81.6
Perceptio 2026.03		81.5
LLaVA-OneVision-7B 2024.08		81.4
Qwen3-VL-8B 2025.11		81.3
IXC2-VL 2024.04		81.2
InternLM-XComposer2 2024.06		81.2
Claude-3.5-Sonnet 2025.06		81.2
Qwen3-VL 2025.12		81
IXC2-4KHD 2024.04		80.9
POINTS-7B 2024.09		80.9
NoisyRollout-7B 2026.02		80.73
Penguin-VL 2026.03		80.7
InternVL-Chat-v1.5 2024.06		80.6
SAYO-Qwen-4B 2026.02		80.51
Gemini-1.5 2026.02		80.3
Claude3.5-Sonnet 2024.12		80.2
Claude3.5-Sonnet 2024.09		80.2
NVLM-D-72B 2024.12		80.1
SAP 2026.02		80.1
Sa2VA 2026.03		79.9
ViGoRL 2026.02		79.7
Cambrian-34B 2024.12		79.5
Phantom-7B 2024.12		79.5
Cambrian-34B 2024.09		79.5
LLaDA-o 2026.03		79.3
Qwen3-VL-8B-Instruct 2026.02		79.2
Gemini-1.5-Pro 2024.12		79.1
Gemini-1.5-Pro 2024.09		79.1
Gemini-1.5-Pro 2025.06		79.1
SenseNova-SI 2025.11		79
InternVL2-4B 2024.12		78.9

Showing 100 of 247 rows