OpenVLM AI2D - Life Cycles: leaderboard

OpenCompass OpenVLM AI2D subcategory score for diagram questions about life cycles.

Metric: Accuracy (%). Source: huggingface.co. Saturation forecast: Estimated already saturated. 284 models tracked.

Top models

#ModelScoreOverall rank
1Gemini 2.5 Pro90.6#152
2InternVL3-78B89.4#352
3GPT-589.2#98
4InternVL3-38B89#405
5GPT-4o ChatGPT87.5
6GPT-5 Mini (2025-08-07)87.3#172
7GPT-4.587.3#274
8GPT-4.1 (2025-04-14)86.8#250
9Qwen 2 VL 72B86.8#460
10InternVL2.5-78B86.1#431
11Qwen 2.5 VL 7B84.9#643
12InternVL3-8B84.2#620
13Gemini 2.0 Flash83.7#342
14InternVL3-14B83.7#505
15Gemini 1.5 Flash (002)83.2#583

No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.

Interactive version: theaggregate.ai/benchmark?slug=openvlm-ai2d-life-cycles · How It Works · Data refreshed daily, snapshot 2026-10-20.