You'll noticeTier: how much it should matter to you
4Useful: my rating, 1 to 5
5Signal: worth watching, 1 to 5
ModelsArea: what it touches
ImprovementsKind: in v2.1.295,
What probably matters to youSection of the release
Unclear It is not stated which feature uses this list or whether a remote setting controls it.
What
A list of models in Claude Code now includes an entry for claude-haiku-5-5 with its thinking set to medium effort. Thinking is the model's step-by-step reasoning before it answers, and effort controls how much reasoning it does. The entry sits next to an existing claude-fable-5 entry, which uses high effort.
The model name claude-haiku-5-5 was already in Claude Code, but it now appears in noticeably more places.
Why
Wherever this list is used, Haiku 5.5 now starts from a medium level of reasoning rather than having no effort setting.
