You'll noticeTier: how much it should matter to you
2Useful: my rating, 1 to 5
3Signal: worth watching, 1 to 5
EffortArea: what it touches
ImprovementsKind: in v2.1.317,
ImprovementsSection of the release
Unclear Whether the reworked checks change when xhigh effort is actually offered to you is not settled.
What
The effort level controls how much Claude thinks before it answers. On each turn (one round of you asking and Claude answering), Claude Code checks whether the xhigh effort level is allowed. That check has been split into two separate checks, and one of them adds extra conditions.
When the model you pick does not support the effort level set for the session, Claude Code now resets the session's effort level. Usage reporting also gains a field that records when typing ultrathink raised the effort level.
Why
This can change when ultrathink or xhigh effort is available to you. If you switch to a model that cannot run at the session's current effort level, the level is set back rather than left in place.
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 doubtWhether the reworked checks change when xhigh effort is actually offered to you is not settled.
