GPT-5
Code generation regressions and unreliable output quality
8Post-update Codex exhibits significant regressions in previously stable workflows, generating code with logical inconsistencies, ignoring design specifications (e.g., front-end ignoring provided UI designs), and requiring multiple re-runs and manual fixes.
High API costs for flagship models at scale
5Developers face high costs when using flagship OpenAI models like GPT-5, especially at high volume usage, making cost management a significant concern for production applications.
Increased refusals and over-cautious behavior in GPT-5.x
5ChatGPT's GPT-5.x models decline requests at a higher frequency than previously, citing safety concerns for benign queries. Creative writing, hypothetical scenarios, and technical troubleshooting prompts trigger refusals that did not occur a year ago. Iterative RLHF tuning has made the model progressively more conservative.
GPT-5 performance degradation on simple tasks
4GPT-5 can feel slower than GPT-4o for simpler, everyday queries and coding tasks. Community backlash occurred regarding performance degradation for simple coding tasks before OpenAI fine-tuned model routing.
Stricter Message Limits and 'AI Shrinkflation'
4GPT-5 launch introduced stricter message limits for paid subscribers and removed access to older, preferred models, creating a perception of 'AI shrinkflation' where users receive fewer capabilities for the same cost.