Updated
Claude Sonnet 5.5 vs GPT-6.1 Sol
Claude Sonnet 5.5 had 32% fewer counted bugs per line than GPT-6.1 Sol (0.96 vs 1.41).
| Claude Sonnet 5.5 | GPT-6.1 Sol | |
|---|---|---|
| Bug rate | ||
| Relative bug rate | 0.96 | 1.41 |
| Rank | #2 of 8 | #6 of 8 |
| By bug type | ||
| Logic bugs(statistically tied) | 0.74 | 0.87 |
| Testing and docs | 1.29 | 2.77 |
| By language | ||
| TypeScript | 0.54 | 0.99 |
| Python | 1.70 | 0.75 |
| Details | ||
| Provider | Anthropic | OpenAI |
| Released | Sep 28, 2026 | Sep 29, 2026 |
| List price | $2 input / $10 output per 1M tokens | $2 input / $10 output per 1M tokens |
| Share of AI-written code, week of Sep 28 | 2.4% | 4.7% |
1.00 = average · lower is better · = statistically tied
More comparisons
Questions and answers
- Which has fewer bugs, Claude Sonnet 5.5 or GPT-6.1 Sol?
- Claude Sonnet 5.5 had 32% fewer counted bugs per line than GPT-6.1 Sol (0.96 vs 1.41). Claude Sonnet 5.5 ranks #2 of 8 and GPT-6.1 Sol #6.
- Which is cheaper, Claude Sonnet 5.5 or GPT-6.1 Sol?
- Both list at $2 input / $10 output per 1M tokens.
- Does it depend on the programming language?
- In TypeScript, Claude Sonnet 5.5 had the lower rate (0.54 vs 0.99); in Python, GPT-6.1 Sol had the lower rate (0.75 vs 1.70).
- Does it depend on the kind of bug?
- For logic bugs, they were statistically tied (0.74 and 0.87); for testing and documentation issues, Claude Sonnet 5.5 had the lower rate (1.29 vs 2.77).