User: Astra Has 'Hemorrhaged IQ' in Codex, Opus 5.5 Is King Again
sterlingcrispin · x · 2026-09-26
Artist Sterling Crispin reports a reversal in model quality: a week after ditching Claude for feeling dumber, he finds Astra inside Codex has 'hemorrhaged IQ' — responses feel like Abbott and Costello's 'Who's on first?' skit — making Claude Opus 5.5 his top pick again. He wonders what's going on at OpenAI. The post captures how frequently updated frontier models swing in quality week to week.
Related event: Developers Say Astra's Performance in Codex Dropped Sharply Within a Week(2 posts)→
More from Models
- Claude one-shots a nine-loop scattering amplitude, a frontier physics calculation — Dr_Singularity · 2026-09-26
- Rumor roundup: Sonnet 5.5, Kimi K3.1, Gemini 4 and more next-gen models incoming — kimmonismus · 2026-09-26
- Stanford professor: frontier AI models show pretense of understanding in open domains — anshulkundaje · 2026-09-26
- HyperWrite founder Matt Shumer declares 'Local models are useless' — mattshumer_ · 2026-09-26
- Users blast Grok's 'absurd' refusals, SuperGrok subscribers cancelling over censorship — Promptmethus · 2026-09-26
- Every's Vibe Check: Anthropic's Fable 5.1 Is Stronger, Friendlier, and Uses Half the Tokens of Opus 5 — every · 2026-09-26