Interesting shift on the βhonesty about its own progressβ part.
Thatβs where these models actually start becoming more reliable in real workflows. Have you tried Opus 4.8 yet?
Introducing Claude Opus 4.8: it builds on Opus 4.7 with sharper judgment, more honesty about its own progress, and the ability to work independently for longer than its predecessors.
Available today at the same price.