Updated
Which AI model writes the most performance bugs?
As of Sep 27, 2026, Claude Sonnet 5 had the highest rate of performance bugs, 117% above the all-model rate (2.17), statistically tied with GPT-6 Astra. GPT-5.5 had the lowest, 38% below average (0.62), statistically tied with 4 other models.
| Rank | Model | Rate |
|---|---|---|
| 1= (statistically tied) | GPT-5.5 | 0.62 |
| 2= (statistically tied) | Claude Fable 5 | 0.69 |
| 3= (statistically tied) | GPT-5.6 Sol | 0.81 |
| 4= (statistically tied) | Claude Opus 4.8 | 0.92 |
| 5= (statistically tied) | Claude Fable 5.1 | 0.97 |
| 6= (statistically tied) | GPT-6 Astra | 1.26 |
| 7= (statistically tied) | Claude Opus 5 | 1.32 |
| 8= (statistically tied) | Claude Sonnet 5 | 2.17 |
1.00 = all-model rate · lower is better · = statistically tied