AI BriefingNew Claude Features
Key point
llm-anthropic 0.25 supports Claude Opus 4.7 and new thinking options.
Details
llm-anthropic 0.25 has been released, updating support for Anthropic models, especially the Claude family.
The biggest change is support for the new model claude-opus-4.7. This model can use up to xhigh for thinking_effort, and the related issue number is shown as #66.
Additionally, new boolean options thinking_display and thinking_adaptive have been added. However, the summary output of thinking_display is currently only available in JSON output or JSON logs.
The default max_tokens value has also been increased to the maximum allowed by each model. At the same time, the structured-outputs-2025-11-13 beta header used by older models is no longer used.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.
Related
US Appeals Court Upholds Pentagon Blacklisting of Anthropic as Supply Chain Risk