Updated
Compare AI coding models
As of Nov 9, 2026, 7 ranked AI coding models make 21 head-to-head comparisons on relative bug rate.
GPT-6.1 Sol#2 · 0.74
- vs GPT-6 Luna0.31
- vs Claude Fable 5.10.88
- vs GPT-6 Astra0.95
- vs Claude Sonnet 5.51.04
- vs Claude Opus 5.51.11
- vs Grok 4.71.61
Claude Sonnet 5.5#5 · 1.04
- vs GPT-6 Luna0.31
- vs GPT-6.1 Sol0.74
- vs Claude Fable 5.10.88
- vs GPT-6 Astra0.95 (statistically tied)
- vs Claude Opus 5.51.11 (statistically tied)
- vs Grok 4.71.61
GPT-6 Luna#1 · 0.31
- vs GPT-6.1 Sol0.74
- vs Claude Fable 5.10.88
- vs GPT-6 Astra0.95
- vs Claude Sonnet 5.51.04
- vs Claude Opus 5.51.11
- vs Grok 4.71.61
Claude Opus 5.5#6 · 1.11
- vs GPT-6 Luna0.31
- vs GPT-6.1 Sol0.74
- vs Claude Fable 5.10.88
- vs GPT-6 Astra0.95
- vs Claude Sonnet 5.51.04 (statistically tied)
- vs Grok 4.71.61
Grok 4.7#7 · 1.61
GPT-6 Astra#4 · 0.95
- vs GPT-6 Luna0.31
- vs GPT-6.1 Sol0.74
- vs Claude Fable 5.10.88 (statistically tied)
- vs Claude Sonnet 5.51.04 (statistically tied)
- vs Claude Opus 5.51.11
- vs Grok 4.71.61
Claude Fable 5.1#3 · 0.88
- vs GPT-6 Luna0.31
- vs GPT-6.1 Sol0.74
- vs GPT-6 Astra0.95 (statistically tied)
- vs Claude Sonnet 5.51.04
- vs Claude Opus 5.51.11
- vs Grok 4.71.61
1.00 = average · lower is better · = statistically tied