Claude Sonnet 5.5 vs GPT-6.1 Sol
GPT-6.1 Sol rates 154 points higher; the 95% intervals do not overlap.
Rating
- Claude Sonnet 5.5953±22
- GPT-6.1 Sol1107±17
GPT-6.1 Sol rates 154 points higher; the 95% intervals do not overlap.
| Claude Sonnet 5.5 | GPT-6.1 Sol | |
|---|---|---|
| Leaderboard rank | 22 of 26 | 3 of 26 |
| Win rate | 27.8% | 41.8% |
| Completion | 85.5% | 81.4% |
| Speed | 0.7 min | 2.9 min |
| Recovery | 84.6%58–96%, n = 13 | 100.0%77–100%, n = 13 |
| Cost per completion | $0.22 | $1.66 |
| Phone rank | 15 of 24 | 8 of 24 |
Under Recovery, its 95% interval and n, in failed actions. Last updated .
In eligible head-to-head comparisons, Claude Sonnet 5.5 has a 18.4% decisive win rate and GPT-6.1 Sol has 59.2%. 22.4% end without a decisive preference. These observed rates do not establish which model is best on every task.
No. Completion for Claude Sonnet 5.5 and GPT-6.1 Sol records whether the agent reported finishing. Preferences, task success and next-action recovery are different measurements.
Read speed and estimated cost alongside completion and preference. An early failure can be fast. Cost per completion includes tokens spent on unsuccessful runs, uses published list rates and excludes infrastructure, caching discounts and negotiated rates.
Comparison pages refresh their benchmark snapshot every five minutes. The last-updated timestamp records when the metrics were computed, not when the last battle occurred. Model rank may move between snapshots.