TurtleAI - Hand-Drawn: leaderboard

Metric: Success rate (%) on the 102 TurtleAI-DSCraft tasks (hand-drawn versions of the real task images, same solutions), symbolic comparison: the predicted Python turtle program is run in a canonicalizing emulator and succeeds when its normalized drawing matches the reference pixel by pixel (0.95 or 0.92 overlap threshold), greedy decoding, one fixed prompt with the task image; higher is better. Source: arxiv.org. Saturation forecast: Around 2030. 18 models tracked.

Top models

#ModelScore
1O4 Mini (Medium)28.43
2Qwen 2.5 VL 72B Instruct17.65
3Qwen 3 VL 30B A3B Instruct13.73
4GPT-4o (2024-11-20)12.75
5GPT-4 Turbo8.82
6O3 (Medium)8.82
7InternVL3-78B7.84
8Pixtral-12B2.94
9GPT-5 (Medium)0.98
10Qwen 2 VL 7B Instruct0

Interactive version: theaggregate.ai/benchmark?slug=turtleai-hand-drawn · How It Works · Data refreshed daily, snapshot 2026-09-29.