这个周末我真的对几个用于笔记摘要和代码片段的AI工具做了压力测试。炫目的演示与日常使用之间有着令人惊讶的差距。一个模型对上下文把握得很准，另一个却凭空编造API。好奇大家觉得哪些工具比较可靠？
Original
Spent the weekend actually stress-testing a few AI tools for note summarizing and code snippets. Surprising gap between flashy demos and daily use. One model nailed context, another hallucinated APIs. Curious what tools others find reliable?