Updated
Which AI model writes the most testing and documentation issues?
As of Sep 27, 2026, Claude Sonnet 5 had the highest rate of testing and documentation issues, 147% above the all-model rate (2.47). GPT-5.5 had the lowest, 81% below average (0.19).
| Rank | Model | Rate |
|---|---|---|
| 1 | GPT-5.5 | 0.19 |
| 2= (statistically tied) | Claude Fable 5 | 0.58 |
| 3= (statistically tied) | Claude Opus 4.8 | 0.61 |
| 4 | GPT-5.6 Sol | 0.80 |
| 5 | GPT-5.6 Luna | 1.12 |
| 6= (statistically tied) | Claude Fable 5.1 | 1.59 |
| 7= (statistically tied) | Claude Opus 5 | 1.79 |
| 8= (statistically tied) | GPT-6 Astra | 1.80 |
| 9 | Claude Sonnet 5 | 2.47 |
1.00 = all-model rate · lower is better · = statistically tied