AI 导读
我越用前沿模型，就越不相信有哪个模型能通吃一切。 - Opus 5：最擅长用生成的 html 教我，但啰嗦又粗糙。 - GPT-5.6 Sol：后端很强，前端偏弱。 - Kimi K3 / GLM-5.2：日常编码 90% 的场景下便宜高效，但在写 kernel 这类 AI 研究任务上较弱。 每家实验室都在为 AGI 训练。但今天的 LLM 仍然是专才。
正文
The more I use frontier models, the less I believe in one model to rule them all.
- Opus 5: best at teaching me with generated html, but verbose and sloppy.
- GPT-5.6 Sol: great backend, weak frontend.
- Kimi K3 / GLM-5.2: cheap and efficient for 90% of day-to-day coding, but weaker at AI research tasks like writing kernels.
Every lab is training for AGI. But today’s LLMs are still specialists.
来源：@Yuchenj_UW · x.com