Anthropic’s tools like Sonnet 4.6 are far ahead of their competition for doing useful work. I use these tools every day and recently created an agentic architecture swapping in models from multiple sources via API calls. Sonnet 4.6 is the boss for accurate rule following and reasoning, but it is pricey on a per call/basis.
Do you use LangGraf or Autogen?