Users Complain About Severe Claude Degradation
-rgo- · reddit · 2026-07-17
The author canceled their Pro subscription and switched back to ChatGPT, citing Claude's latest version's increased "refusal to execute," high error rates, and inexplicable throttling.
They gave examples where Fable 5, Opus, and Sonnet all insisted they lacked the tool capability to access profile memory, only changing their tune when provided with screenshots; even then, the models questioned if the screenshots were AI-generated. The author feels this is like "gaslighting" the user and suspects it's a systemic degradation in recent checkpoints.
Related event: Users Complain About Claude's Declining Speed and Quality(4 posts)→
More from Models
- MissionReport benchmark: only 3 of 540 AI-generated government reports pass all criteria — Exp_Mark · 2026-10-27
- Local beats cloud: Qwen 3.8 FN handles private student data as well as frontier models — dr_lm · 2026-10-27
- Aidan Gomez: model's CoT said no tool calls right before making them — aidangomez · 2026-10-27
- Nat Lambert: Cohere's model is narrow RL training on internal infra ill-suited for open weights — natolambert · 2026-10-27
- OpenAI teases DOTS is coming to Europe 'soon' — mark_k · 2026-10-27
- Clifford Pickover challenges AI to decipher a message he encoded in Norse runes — pickover · 2026-10-27