OpenAI has added new modes to its GPT-5 model with 3000 messages per week limit, while restoring GPT-4o for paid users as it plans further personality refinements for ChatGPT.
CEO Sam Altman took to X earlier today announcing that users can now pick between three modes when using GPT-5 i.e. Auto, Fast, and Thinking. Auto adjusts automatically balancing speed and depth, Fast prioritizes quick responses, while Thinking offers higher reasoning power at the cost of slower replies.
The Thinking mode now supports a generous weekly cap of three thousand messages and a context window reaching 196 000 tokens. After hitting that limit, users can continue in a lighter version called Thinking mini. Altman noted that usage limits may be tweaked over time based on demand and usage patterns.
The updates follow user frustration since GPT-5’s launch, particularly over the sudden disappearance of GPT-4o. Feedback pointed to how much users valued its warmth and companion-like presence. In response OpenAI has restored GPT-4o in the model picker for all paid accounts, promising advance notice should it be retired again.
A new toggle labelled “Show additional models” in ChatGPT’s web settings lets subscribers access options like o3 4.1 and GPT-5 Thinking mini. Meanwhile GPT-4.5 remains available exclusively to Pro users, owing to its high GPU cost.
Altman also confirmed plans to refine GPT-5’s conversational style to make it feel warmer yet less try hard than GPT-4o. The company is working toward more per-user customization so users can shape the AI’s style according to their personal taste.
Updates to ChatGPT:
You can now choose between “Auto”, “Fast”, and “Thinking” for GPT-5. Most users will want Auto, but the additional control will be useful for some people.
Rate limits are now 3,000 messages/week with GPT-5 Thinking, and then extra capacity on GPT-5 Thinking…
— Sam Altman (@sama)
These changes come amid widespread criticism that GPT-5’s model routing delivered inconsistent results and lacked the emotional subtlety users expect. OpenAI says these updates mark its quick effort to respond and reassure users.
Altman highlighted that the goal is to deliver both robust performance and a comfortable experience. By introducing modes and restoring familiar AI voices while planning for customization, OpenAI signals it intends to serve both everyday users and power users effectively.
Want to see more of our stories on Google?
P.S. Want to keep this site truly independent? Support us by buying us a beer, treating us to a coffee, or shopping through Amazon here. Links in this post are affiliate links, so we earn a tiny commission at no charge to you. Thanks for supporting independent Canadian media!
