Anthropic 关于 Claude Sonnet 5.5 的客户证言
“在 Epic 公司的早期测试中，Claude Sonnet 5.5 达到了人们对其高阶模型所预期的质量标准，在系统设计审计和数据流审查中均表现稳健。该新模型可管理数万行游戏系统架构代码，响应迅捷，能处理持续数小时的复杂任务，且所需指令性提示更少。”
原始摘录
“In Epic’s early testing, Claude Sonnet 5.5 cleared the same quality bar you’d expect from a higher-tier model, holding up on a system design audit and a data flow review. The new model managed tens of thousands of lines of code for gameplay system architecture, kept responses snappy, handled multi-hour tasks, and delivered with less prescriptive prompting.”