Effortless Caching 💾

New model releases and Anthropic fixed cache busting when changing effort levels mid-task

• Claude Opus 5.5
  After some testing and working with it all day, Opus 5.5 is really good! It’s cheaper, seems more consistent and fixed the god awful writing that made me question my ability to speak English.

• GPT 6 Sol and Luna
  OpenAI follows up their GPT 6 line with their smaller models. Nonetheless Sol still is great value for your money and Luna might be useful for small cheap tasks.

• Getting the most out of Opus 5.5 in Claude and Claude Code
  Addy Osmani gives practical advice on how Anthropic expects you to work with their newest model.

• Using Claude Code: Spending your effort
  Anthropic fixed cache busting when you are switching effort levels mid-task, yay! This also means that there are now real use-cases for switching effort level while working.

• 10 tells of a slop ui
  Don’t produce tasteless slop. If you ever wondered why “vibe-coded” products look and feel all the same, here are some tells of slop ui