Premium

ChatGPT gets Auto, Fast, and Thinking options, but its limited to paid GPT-5 users

OpenAI CEO Sam Altman says ChatGPT users can now choose between "Auto", "Fast" and "Thinking" modes, but it looks like the new feature is limited to paid users.

ChatGPT no longer lets free users switch to legacy models like GPT-4o. (Express Photo)

A few days after OpenAI unveiled GPT-5, its fastest and most capable AI model to date, several users asked the company to bring back GPT-4o, with many paid users even cancelling their ChatGPT subscription.

Following this, OpenAI CEO Sam Altman said that they will be bringing a wide range of upgrades to the AI chatbot. Now, in a post on X, Altman has announced that ChatGPT users can now choose between “Auto”, “Fast” and “Thinking” options when using GPT-5.

To give you a quick recap, GPT-5 came with a new real-time router which automatically determined which model to use for a particular task. The approach was good on paper, but Altman himself had confirmed that ChatGPT’s performance was worse than before and said that the company would be more transparent about which model the AI chatbot is using to answer a query.

Also Read | Gemini Live now connects to Keep, Tasks, Calendar and Maps: Here’s how to check

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) August 13, 2025

Altman went on to say that the Auto mode will suffice for most users, but the other options might come in handy for people who want more control over ChatGPT’s responses.

Apart from the new mode selection options, OpenAI has also increased the rate limits to 3,000 messages per week with GPT-5 Thinking, which has a context limit of 196,000 tokens. If you happen to exhaust it, ChatGPT will use GPT-5 Thinking mini. The AI startup also said that they might change the rate limit depending on usage.

Sam Altman also talked about GPT-5’s ‘cold’ personality, saying that the new model “should feel warmer than the current personality but not as annoying (to most users) as GPT-4o.”

And while ChatGPT Plus, Pro, Teams and Enterprise users now have the option to switch back to GPT-4o, it looks like the company won’t be making the feature available for those on the free tier.