Skip to content

Benchmark

AI2D

Vision benchmarks measure how well models understand images. Scores are percentages.

Models measured
69
Scores recorded
71
Sources
2
IQ weight
1.11

Score distribution

82.4%
Mean
84.2%
Median
10%
Std dev
70.7%
10th percentile
92.7%
90th percentile
25.6%
Min
25.6%94.7%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.11 for chat models, measured over 68 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
1Claude 3.5 Sonnet· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim94.7%
2Qwen3.6 Plus· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim94.4%
3GPT 4o· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim94.2%
4Pixtral Large· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim93.8%
5Qwen3.5 122B-A10B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim93.3%
6Hunyuan Standard Vision· OpenVLM Leaderboard data93.2%
7Mistral Small 3.2 24B Instruct2506· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim92.9%
8Qwen3.5 27B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim92.9%
9Qwen3.6 35B A3B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim92.7%
10Qwen3.5 35B A3B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim92.6%
11Llama 3.2 11B Instruct· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim91.1%
12InternVL3 78B· OpenVLM Leaderboard data89.8%
13Qwen3 VL 235B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it.89.7%
14Gemini 2.5 Pro· OpenVLM Leaderboard data89.5%
15GPT 5· OpenVLM Leaderboard data89.5%
16Qwen3 VL 32B Instruct· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it.89.5%
17Qwen3 VL 235Bthinking· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it.89.2%
18SenseNova-V6-Pro· OpenVLM Leaderboard data89.2%
19Doubao Seed 1.8· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim89.1%
20Qwen3 VL 32B Thinking· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it.88.9%
21internvl3-38b· OpenVLM Leaderboard data88.7%
22Qwen2.5 VL 72B Instruct· OpenVLM Leaderboard data
Values from 2 sources
88.5%
23Qwen2-VL-72B· OpenVLM Leaderboard data88.3%
24GPT-4.5 Preview· OpenVLM Leaderboard data87.2%
25qwen3-vl-30bthinking· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it.86.9%
26GPT-5 mini· OpenVLM Leaderboard data86.7%
27ChatGPT-4o Latest· OpenVLM Leaderboard data86.3%
28GPT 4.1· OpenVLM Leaderboard data85.9%
29Qwen3 VL 8B Instruct· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it.85.7%
30GLM 4 Vision Plus· OpenVLM Leaderboard data85.1%
31internvl3-8b· OpenVLM Leaderboard data85.1%
32qwen3-vl-30b· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it.85%
33Qwen: Qwen3 VL 8B Thinking· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it.84.9%
34Gemma 3 27B IT· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim
Values from 2 sources
84.5%
35InternVL2 26B· OpenVLM Leaderboard data84.5%
36gemini-1.5-flash-002· OpenVLM Leaderboard data84.4%
37Qwen2.5-VL 7B Instruct· OpenVLM Leaderboard data84.3%
38Gemma 3 12B IT· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim
Values from 2 sources
84.2%
39DeepSeek VL2· OpenVLM Leaderboard data
Values from 2 sources
83.8%
40Molmo 72B· OpenVLM Leaderboard data83.4%
41gemini-1.5-pro-002· OpenVLM Leaderboard data83.3%
42Qwen2.5-Omni 7B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim83.2%
43Gemini 2.0 Flash· OpenVLM Leaderboard data83.1%
44Phi 4 Multimodal Instruct· OpenVLM Leaderboard data
Values from 2 sources
83%
45MiniCPM-V 2.6· OpenVLM Leaderboard data82.1%
46Moonshot V1 8K· OpenVLM Leaderboard data81.8%
47GPT-5 nano· OpenVLM Leaderboard data81.4%
48qwen2.5-vl-3b-instruct· OpenVLM Leaderboard data81.4%
49Grok 2 Vision 1212· OpenVLM Leaderboard data81.2%
50Molmo 7B-D· OpenVLM Leaderboard data81%

Showing 1–50 of 71 results.

Data credits: OpenVLM Leaderboard (OpenCompass / VLMEvalKit), Apache-2.0 (licence) · Data by LLM Stats