llm-anthropic exposes Claude reasoning traces by default
Simon Willison's llm-anthropic library now surfaces reasoning traces by default for Claude models that generate them, a shift that exposes intermediate problem-solving steps to developers building on Anthropic's stack. The update also introduces structured error handling via a new ClaudeRefusal exception, letting applications distinguish between model refusals and other failures. This matters because reasoning transparency is becoming table stakes for LLM tooling; developers need visibility into model behavior for debugging, trust, and compliance. The exception handling improvement reduces boilerplate in production systems.
Modelwire context
ExplainerThe llm-anthropic update surfaces Claude's internal reasoning steps by default, making the model's problem-solving process visible to developers rather than hidden inside a black box. This is distinct from simply logging API calls; it exposes intermediate thinking that was previously inaccessible.
This release is part of a coordinated observability push across Simon Willison's tooling ecosystem. The llm CLI reached 0.34 just yesterday with response timing metrics, and now llm-anthropic adds reasoning transparency. Together, these incremental updates reflect a pattern: as Claude Fable 5.1 ships with better coding and research capabilities, the developer experience around those models is maturing in parallel. Visibility into both latency and reasoning chains directly supports the agentic workflows that Anthropic is now pricing aggressively to capture. The tooling layer is catching up to the capability layer.
If other LLM client libraries (OpenAI's Python SDK, Anthropic's direct SDK) ship similar reasoning trace defaults within the next two quarters, that confirms reasoning transparency is becoming table stakes. If they don't, llm-anthropic remains a niche advantage for developers already in Willison's ecosystem.
Coverage we drew on
- llm 0.34 · Simon Willison
This analysis is generated by Modelwire’s editorial layer from our archive and the summary above. It is not a substitute for the original reporting. How we write it.
MentionsSimon Willison · Anthropic · Claude · llm-anthropic · Claude Fable 5.1
Modelwire Editorial
This synthesis and analysis was prepared by the Modelwire editorial team. We use advanced language models to read, ground, and connect the day’s most significant AI developments, providing original strategic context that helps practitioners and leaders stay ahead of the frontier.
Modelwire summarizes, we don’t republish. Simon Willison originally reported this story as “llm-anthropic 0.28”. The full content lives on simonwillison.net. If you’re a publisher and want a different summarization policy for your work, see our takedown page.