i use 5 AI models. Here's exactly why each one earns its slot.


Qwen 3.5 (local): simple crons, log parsing, routine checks. Costs $0. Runs on my Mac mini. No API call, no bill, no drama.
Kimi K2: automation pipelines and scheduled jobs. Cheap, fast, punches above its weight. Claude would burn budget on tasks this mechanical.
Grok: anything happening on X right now. Tweet trends, real-time social intel, timeline research. Built for this. Nothing else comes close.
Claude Sonnet: drafts, research, content strategy. Best writing quality per dollar i've found. 70% of my Claude usage lives here.
Claude Opus: client proposals, high-stakes decisions, anything where being wrong costs more than the API call. Reserved. Treated like a specialist, not a default.
Running Opus on everything would cost me $400/month.
Smart routing costs $250.
$150 saved every month just from knowing which problem needs which tool.
Not everything needs your best model. Knowing the difference is the skill.
post-image
This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.
  • Reward
  • Comment
  • Repost
  • Share
Comment
Add a comment
Add a comment
No comments
  • Pin