Max plan 到底有多划算
切到 API 计费后终于看到了 Claude Code session 的真实成本。笔记本 14 天烧了几千美元，云端 devbox 又烧了几千。Max plan 是固定月费。这笔账太离谱了。
- AI (190)
- Agents (110)
- Foundry (2)
- Claude Code (31)
- Easter Eggs (2)
- Fun (2)
- Software Development (24)
- Frontend (4)
- Agent Runtimes (3)
- 自动化 (2)
- macOS (4)
- Automation (2)
- Compute (4)
- Philosophy (2)
- 微信 (1)
- WeChat (1)
- Chips (2)
- Hardware (2)
- GPU (2)
- TPU (2)
- Mio (66)
- Lumi (8)
- Writing (8)
- NLP (6)
- Product (20)
- Engineering (18)
- Chinese (2)
- Design (8)
- Voice (8)
- DevOps (8)
- Linux (8)
- VPN (4)
- 翻墙 (2)
- 科学上网 (2)
- Self-hosting (2)
- Career (4)
- Economics (4)
- China (3)
- Psychology (3)
- Labor (2)
- Memory (4)
- Claude (4)
- LLM (11)
- GPT (2)
- Business (2)
- OpenClaw (8)
- TTS (2)
- GCP (4)
- Workflow (2)
- Ops (2)
- Cost Optimization (2)
- Agent Teams (4)
- PanPanMao (4)
- Best Practices (1)
Claude Code Max: The Most Lopsided Deal in AI
I switched to API billing on Claude Code and finally saw the real cost of my sessions. Thousands of dollars on my laptop in 14 days. Thousands more on my cloud devbox. The Max plan is a flat monthly fee. The math is absurd.
Benchmark 分数高又怎样
GPT 5.4 在各项 benchmark 上全面领先。但当我把同一个复杂的产品战略问题扔给两个模型时，benchmark 分数和真实输出质量之间的鸿沟令人震惊。
GPT 5.4 vs Opus 4.6: Why Benchmarks Stopped Mattering
GPT 5.4 dominates every benchmark. But when I gave both models the same complex product strategy prompt, the gap between benchmark scores and real-world output was staggering. Here's what actually happened.