📄 中文摘要
在对比 Claude Sonnet 4.6 和 Gemini 3.1 Pro 的使用体验中，Claude 在处理复杂的调试任务时表现出色，能够全面理解提示并准确捕捉所有约束条件，给人一种与资深开发者合作的感觉。相较之下，Gemini 在算法和逻辑方面也很强，但有时会添加未提及的内容。两者在实际项目中的表现差异显著，Claude 更加专注于用户的具体需求，而 Gemini 则可能会偏离主题。
我连续两周使用了 Claude Sonnet 4.6 和 Gemini 3.1 Pro。这里是没人告诉你的事情。
📄 English Summary
I used both Claude Sonnet 4.6 and Gemini 3.1 Pro for two weeks straight. Here's what nobody tells you.
Using Claude Sonnet 4.6 and Gemini 3.1 Pro for two weeks revealed distinct differences in their capabilities. Claude excels at understanding complex debugging tasks, catching all constraints and providing focused, relevant explanations, akin to working with a senior developer. In contrast, Gemini demonstrates strong algorithmic and logical skills but occasionally adds unnecessary information not mentioned in the prompt. The practical application of both AIs highlights Claude's ability to stay on topic while Gemini may diverge from the user's specific needs.