I called this a few days ago.Claude Sonnet 5 just shipped.It is now the default model for Claude Free and Pro. Anthropic's pitch is clear: agentic performance close to Opus 4.8, at 40% of the API price.Can’t wait to test it.I’ll share some demo results later.

The model everyone was talking about is back.Claude Fable 5 is now live again!For me, what stands out most is how Anthropic handled the whole situation.Over the past few weeks, there has been a lot of debate about AI safety, model access, and how much freedom powerful AI tools should have.Anthropic’s answer seems simple:Bring Fable 5 back.Fix the areas that needed work.Watch how people use it.Improve as they learn.That feels like the right approach.For most people, normal coding work should stil