Updated
Which AI model writes the most testing and documentation issues?
As of Nov 14, 2026, Grok 4.7 had the highest rate of testing and documentation issues (1.62, 62% above the all-model rate), statistically tied with Claude Sonnet 5.5. GPT-6 Luna had the lowest (0.44, 56% below average).
| Rank | Model | Rate |
|---|---|---|
| 1 | GPT-6 Luna | 0.44 |
| 2= (statistically tied) | GPT-6.1 Sol | 0.93 |
| 3= (statistically tied) | Claude Fable 5.1 | 0.93 |
| 4= (statistically tied) | GPT-6 Astra | 1.17 |
| 5= (statistically tied) | Claude Opus 5.5 | 1.24 |
| 6= (statistically tied) | Claude Sonnet 5.5 | 1.24 |
| 7= (statistically tied) | Grok 4.7 | 1.62 |
1.00 = all-model rate · lower is better · = statistically tied