Plus Users Complain About Instability on Complex Tasks
Nuwen-Pham · reddit · 2026-07-12
A ChatGPT Plus subscriber expressed growing doubts about the subscription's value for complex technical work.
The main issues listed include:
- Making errors when processing long diagnostic logs or multi-stage infrastructure fixes
- Unstable file upload processing
- Poor ability to maintain project context and reason across multiple turns
- Low reliability when generating executable code
- Forgetting confirmed facts, fabricating details, or outputting environment-incompatible commands
The author suggests ChatGPT is better suited for:
- Researching and comparing solutions
- Summarizing public sources
- Synthesizing information and generating alternative perspectives
- Drafting an initial structure for human review
Since they already possess local inference capabilities for summarization, private document analysis, continuous context, and iterative refinement, they are considering whether the "free version of ChatGPT + local models" sufficiently covers their needs. Finally, they asked other Plus users if they experience similar reliability issues, whether the main benefit of Plus is just increased quotas and features, if there are stable workflows to prevent context drift, and if anyone has downgraded without missing much.
More from Models
- Claim says Kimi was distilled from Fable, sparking a model-attribution jab — cephaloform · 2026-07-22
- Gary Marcus says LLM math skills are like knowing only a car’s engine size — GaryMarcus · 2026-07-22
- OpenAI’s Codex + GPT-5.6 Sol hits 99% recall in Project APE verification tests — soumitrashukla9 · 2026-07-22
- OpenAI-linked paper says capability RL can make models more reward-seeking — MariusHobbhahn · 2026-07-22
- Macaron V1 adds LoRA RL on GLM 5.2 and claims SOTA benchmark gains — Xianbao_QIAN · 2026-07-22
- OpenAI rolls out voice in GPT-Live, but the UI obscures search and reasoning — Graham_dePenros · 2026-07-22