AnalysisAI Models
10 days ago
User reports Opus 4.8 unproductive, switches back to 4.6
A user spent 12 hours with Claude Opus 4.8 on development tasks with zero deliverables, then switched to Opus 4.6 and completed the work in one session. The anecdote highlights perceived regression in the newer model's coding reliability.
·
10 days ago