GPT-5 just got a big new upgrade, and Sam Altman has fixed Plus users' biggest complaint

• GPT-5 now has multiple thinking modes to choose from
• Sam Altman has also confirmed an increase in rate limits for ChatGPT Plus subscribers
• GPT-5 launched less than a week ago to widespread criticism

GPT-5 just got its first major change, and now users can select between different modes when using the new model in ChatGPT.

Confirmed by OpenAI CEO, Sam Altman, on X earlier today, ChatGPT users can now choose between Auto, Fast, Thinking, and Thinking-mini when using GPT-5.

Each new mode offers a different way for GPT-5 to, you guessed it, think. "Auto" lets GPT-5 decide for itself how long to think, Fast" gives you instant answers, "Thinking-mini" thinks quickly, and "Thinking" will take longer to think for better answers.

The change comes following mass backlash related to GPT-5's performance, and will now give users multiple tiers of performance to choose from. We've yet to test all of the new thinking modes; however, when OpenAI decided to limit choice and remove legacy models, the lack of variety was met with widespread criticism.

OpenAI has since reverted back on those decisions, making 4o available again for paid subscribers, and adding the choice of multiple thinking abilities in GPT-5 only further cements the U-turn.

3000 messages a week? Yes please

New thinking modes aren't the only changes coming to GPT-5. Altman also announced the increase in rate limits for the brand new AI model following discontent from ChatGPT Plus users who pay $20/£20 a month to access the premium tier.

At launch, GPT-5's Thinking model was limited to 200 messages per week for Plus subscribers, now Altman says the rate limits have been increased to 3,000 a week. He also notes, "Context limit for GPT-5 Thinking is 196k tokens. We may have to update rate limits over time depending on usage."

Earlier this week, Altman said ChatGPT-5 Pro might be coming to Plus subscribers too, although he now appears to have backtracked, claiming, "we do not have the compute to do it right now."

GPT-5 hasn't even been out a week yet, but OpenAI has started to right the wrongs of the initial launch. With new rate limits and more choices in how long the AI model takes to respond with less or more thinking process, the company is trying to recapture its user base's trust.

You might also like

• GPT-5 free tier – 3 cool features you can try now for nothing
• GPT-5 is here – 5 things you need to know about OpenAI’s ‘most useful’ model yet
• Yes, Sam Altman will be played by Andrew Garfield in 'Artificial', but all we want to know is who’s playing Elon Musk