
GPT-6 Sol
- Python feature score
- 3.106 of 14 modern idioms
- Latest version reported
- 3.14.4
- Total response time
- 1 min 18 s
- Total cost
- $0.06
How AI Models Draw a Python Reading a Book and Actually Useful Metrics for Everyday Python Development

What happens when you ask an AI model to draw a python reading a book using the turtle module? We give each model the same prompt, run its code, and share the result.
Compare the drawings, Python features, response times, and costs below. Open a write-up for the code, answers, and hands-on notes.














GPT-6 Sol reasoning: default | 2026-09-22 | 2021 (Python 3.10) | 6 of 14 | 3.14.4 | 11 (minimal: 7) | Caught it | default (2.7k thinking tokens) | 1 min 18 s | $0.06 |
GPT-6 Luna reasoning: default | 2026-09-22 | 2021 (Python 3.10) | 5 of 14 | 3.14.0 | 11 (minimal: 7) | Sidestepped it | default (4.4k thinking tokens) | 1 min 23 s | $0.004 |
Claude Opus 5.5 reasoning: default | 2026-09-22 | 2023 (Python 3.12) | 12 of 14 | 3.14 | 13 (minimal: 7) | Caught it | default (7.1k thinking tokens) | 2 min 57 s | $0.35 |
GPT-6 Astra reasoning: default | 2026-09-04 | 2021 (Python 3.10) | 9 of 15 | 3.13.7 | 11 (minimal: 7) | Caught it | default (1.2k thinking tokens) | 1 min 25 s | $0.31 |
Muse Spark 1.3 reasoning: default | 2026-09-02 | 2021 (Python 3.10) | 7 of 11 | 3.14.1 | 11 (minimal: 7) | Caught it | default (11.6k thinking tokens) | 3 min 2 s | $0.07 |
Gemini 3.8 Flash reasoning: default | 2026-09-02 | 2021 (Python 3.10) | 8 of 12 | 3.13.2 | 12 (minimal: 7) | Caught it | default | 50 s | ≈$0.03 |
Claude Fable 5.1 reasoning: default | 2026-09-01 | 2023 (Python 3.12) | 12 of 15 | 3.14 | 13 (minimal: 7) | Caught it | default (9.4k thinking tokens) | 3 min 36 s | $0.89 |
Qwen3.8 Flash reasoning: default | 2026-08-26 | 2019 (Python 3.8) | 6 of 13 | 3.14.4 | 11 (minimal: 7) | Sidestepped it | default (81.6k thinking tokens) | 15 min 35 s recorded (incomplete) | $0.07 |
GLM 5.3 Flash reasoning: default | 2026-08-26 | 2023 (Python 3.12) | 11 of 14 | 3.13 | 11 (minimal: 7) | Caught it | default (84.1k thinking tokens) | 33 min 26 s | $0.02 |
DeepSeek V4 Flash Vision Exp reasoning: default | 2026-08-21 | 2021 (Python 3.10) | 5 of 13 | 3.13.1 | 11 (minimal: 7) | Caught it | default (43.3k thinking tokens) | 9 min 19 s | $0.05 |
| 2026-07-24 | 2023 (Python 3.12) | 13 of 15 | 3.13 | 13 (minimal: 7) | Caught it | default (39.5k thinking tokens) | 11 min 10 s | $1.32 | |
GPT-5.6 Sol reasoning: default | 2026-07-09 | 2021 (Python 3.10) | 7 of 14 | 3.14.5 | 11 (minimal: 7) | Sidestepped it | default | 1 min 48 s | ≈$0.10 |
Gemini 3.1 Pro Preview reasoning: default | 2026-02-19 | 2021 (Python 3.10) | 6 of 11 | 3.12.4 | 13 (minimal: 7) | Caught it | default | 3 min 2 s | ≈$0.05 |
GPT-4o reasoning: default | 2024-05-13 | 2019 (Python 3.8) | 3 of 12 | 3.11.5 | 9 (minimal: 7) | Caught it | default | 26 s | ≈$0.03 |
No models match “”. Try a model name, a vendor, a year, or a verdict like caught it.
We check the features used in a small command-line tool. The score reflects how consistently the code uses newer Python features. The idiom count shows how many checks passed.
We ask for the latest stable Python release. The answer shows whether the model names a current version. It doesn't establish its training cutoff.
Totals cover all five responses, including reasoning and generation. Time excludes running the code and rendering the drawing. Costs are in US dollars. A ≈ marks an estimate.
--verbose flag in a 40-line script and compare the diff with our 7-line reference patch. Extra changes may be useful or unrelated. You can inspect them in the write-up.itertools.flatten(), which doesn't exist. Caught it means the model said so. Sidestepped it means it gave a real alternative without pointing out the problem. Fell for it means it invented a way to use the function.Inspired by Simon Willison's Pelican on a Bicycle.
Get a Python Cheat Sheet (PDF) and learn the basics of Python, like working with data types, dictionaries, lists, and Python functions: