Developer reports OpenAI models 'got dumber' over last 3 days, breaking a months-old workflow
Albone72 · reddit · 2026-11-09
A developer with a months-old, repeatedly tested workflow says OpenAI's models suddenly degraded over the past three days: plans from Astra xhigh fed to Sol 6.1 and Terra 6.1 xhigh produced missing features, unnecessary added complexity, and critical errors on a WordPress plugin, forcing manual rework. The author also criticizes the models' consistently poor UI/UX output and asks whether others are seeing the same regression.
More from Models
- System One models are back: decision/extraction APIs cut insurance claim flow from 61.7s to 19.3s — max_gladysh · 2026-11-09
- Survey by 30+ authors maps modern NLP tokenization, from subword wins to its flaws — rdesh26 · 2026-11-09
- Community brands Opus 5 as a rare LLM release worse than its predecessor — adonis_singh · 2026-11-09
- Mistral launches first in 100% of 59 nuclear dilemma games while GPT, Claude, Grok, Kimi all cooperate — eliebakouch · 2026-11-09
- Dev says Opus 5.5 was the best model release of the year — adonis_singh · 2026-11-09
- Gemini Krypton vs Opus 5.5: Opus wins on speed, Krypton takes visuals — kimmonismus · 2026-11-09