Livenerf: Has Opus 5.5 been nerfed yet?(github.com)
858 points by bryan0 | 367 comments
GitHub repository for livenerf, a benchmark that tracks AI model capability over time after release to detect whether models like Claude Opus have been nerfed.
#model-degradation on Hacker News
3 stories tagged model-degradation, newest first.
Fable 5 – Median thinking declined in August(twitter.com)
368 points by espeed | 253 comments
An X post by Lon Lundgren with a chart claiming Anthropic's Fable 5 used 22–51% fewer median thinking tokens in August than July on subscription plans.
Claude Code daily benchmarks for degradation tracking(marginlab.ai)
720 points by qwesr123 | 330 comments
Margin Lab tracker that runs daily SWE-Bench-Pro benchmarks on Claude Code with Opus 4.5 to statistically detect performance degradation.