You'll noticeTier: how much it should matter to you
2Useful: my rating, 1 to 5
3Signal: worth watching, 1 to 5
ModelsArea: what it touches
ImprovementsKind: in v2.1.281,
ImprovementsSection of the release
What
"Thinking" is the step where the model reasons before it answers. Some models are marked as not allowing thinking to be disabled (rejects_disabled_thinking). For those models, Claude Code now always keeps thinking on, whatever the setting says.
In /config, 'Thinking mode' shows the value actually in effect and appears locked. If you try to turn it off, you get the error "Thinking can't be turned off for this model".
Why
Before, the setting could say thinking was off even though that model cannot run that way. Now the setting shows what really happens, and you get a clear message instead of a setting that silently does nothing.
How sure we are
Something disagreesSomething we can check disagrees with this entry, or the writer said they could not settle it.
The writer flagged doubtThe finding does not name which models carry the `rejects_disabled_thinking` capability.
Anthropic's release notes agreeFixed Alt+T and /config offering to turn thinking off on models that can't; thinking now stays on there, with a one-line reason in place of…
