Just a few days after OpenAI launched GPT-5, their quickest and most powerful AI model yet, a bunch of users started requesting the return of GPT-4o. Some paid users even went as far as cancelling their ChatGPT subscriptions.

Recently, in a post on X, Altman revealed that ChatGPT users now have the option to select between “Auto”, “Fast”, and “Thinking” modes when using GPT-5.

GPT-5 introduced a new real time router that figured out which model to use for specific tasks. It sounded great in theory, but Altman admitted that ChatGPT’s performance actually declined compared to earlier versions.

He also mentioned that the company plans to be clearer about which model the AI chatbot is using to respond to questions.

Altman mentioned that the Auto mode should be enough for most users, but the other settings could be useful for those looking for more control over ChatGPT’s replies.

In addition to the fresh mode selection options, OpenAI has upped the rate limits to 3,000 messages weekly with GPT-5 Thinking, which can handle a context limit of 196,000 tokens. If you run out, ChatGPT will switch to GPT-5 Thinking mini.

The AI startup also mentioned that they could adjust the rate limit based on how it’s being used.

Sam Altman mentioned that GPT-5 has a ‘cold’ vibe, explaining that the new version “needs to come across as friendlier than the existing personality but not as bothersome as GPT-4o”.

Additionally, while ChatGPT Plus, Pro, Teams, and Enterprise users can now revert to GPT-4o, it seems that this option won’t be offered to those using the free version.

Subscribe My Channel





Discover more from Connect2ConnectOnline

Subscribe now to keep reading and get access to the full archive.

Continue reading