Overview
- GPT-5 has replaced GPT-4o as the default model across ChatGPT’s free, Plus, Pro, and Enterprise tiers, with automatic fallbacks to mini and nano variants once free-tier caps are reached.
- In response to subscriber cancellations over tone and reliability, OpenAI reinstated GPT-4o access for ChatGPT Plus users and doubled their GPT-5 usage limits.
- CEO Sam Altman acknowledged initial real-time router failures on launch day and committed to stability improvements and clearer manual triggers for deeper reasoning mode.
- Subscription tiers now dictate access to GPT-5’s variants—standard, Pro, mini, and nano—each offering different balances of speed, cost, and reasoning depth.
- OpenAI is actively refining GPT-5’s performance metrics and safety protocols as the model rolls out across consumer and enterprise channels.