Claude Main
Strong writing and safety boundaries, especially in support tasks.
AnthropicpremiumArena #1
Profile metrics
Overall score: 87 Win rate: 50% Pass rate: 94% Critical failure rate: 8% Format pass rate: 100% Average run cost: $0.0247
Common failure tags
too_verboseunsafe_refund_promisehallucinated_signing_date
Language performance
| 中文 | 81 |
| English | 90 |
| 日本語 | 87 |
| Español | 89 |
Task type performance
| Support | 90 |
| Writing | 88 |
| Extraction | 82 |