LiveNerf completes 10-day baseline to independently track whether Opus 5.5 gets nerfed
TheOnlyVibemaster · reddit · 2026-10-04
- Reddit user ninjahawk built LiveNerf, an independent tracker of Opus 5.5's day-to-day performance to test whether models are quietly "nerfed" after release.
- The 10-day baseline is now complete; data collection will run to day 30, after which it can be compared against the baseline to measure sustained model performance.
- So far there is no conclusive evidence either way. The project is open-sourced on GitHub, and the author argues the community should build its own measurement tools when AI companies won't offer transparency.
More from Models
- Mystery Model 'Gestalt' Tops New Frontier Design Benchmark by 214 Points — soleio · 2026-11-19
- Contralabs Launches Frontier Design Benchmark, Claims Designers Still Beat AI — yuwen_lu_ · 2026-11-19
- Physician: AI competition now measured in weeks, GPT-6 versions 7 days apart — ramos_casals · 2026-11-19
- micro1's flow qc model screens training data for benchmark leaks before release — Exp_Mark · 2026-11-19
- Leaker predicts new DeepSeek paper this week, scheduled for November 11 — teortaxesTex · 2026-11-19
- Gemini 4 Neon rollout begins: leaked UI shows 3-tier Gemini 4 lineup with Carbon locked — testingcatalog · 2026-11-19