Chat GPT might not be using the model you think — and it’s also hiding others in settings | Tech Radar

Overview

Chat GPT might not be using the model you think — and it’s also hiding others in settings

There's a whole rotating cast of models working in the background

Details

When you purchase through links on our site, we may earn an affiliate commission. Here’s how it works.

Chat GPT now uses a simplified model picker that can automatically switch between different AI models

Users may not realize that different prompts can trigger different underlying models

Additional models and controls are hidden in settings that most never access

Don't worry, you haven't gone crazy, Chat GPT's model picker (at the top of the screen) is looking a little cleaner this week, with fewer model names cluttering up the interface.

Forget model names like 5.4, 4o and o 3, that's a thing of the past! Chat GPT Plus subscribers will now see just three choices, labeled Instant, Thinking, and Pro. However, the suggestion of transparency over complexity isn't quite what it seems.

The update changes the nature of Chat GPT options from a choice of models to more of a broad style request. The actual model used in an answer will more often be decided by Chat GPT based on your prompt complexity and other settings.

Chat GPT’s new settings range from corporate calm to chaotic bestie

Open AI introduces Chat GPT 5.4 Thinking for solving bigger problems

Those factors will affect whether Chat GPT's answers are from a faster, more lightweight model or a more powerful, power-hungry LLM. You might not even be told which one handled your request in the result.

It puts control of Chat GPT a step beyond the user. Selecting a Chat GPT model used to mean just that. Now, selecting any of the three modes might correspond to any of Chat GPT's stable of models, depending on other elements.

You might get answers almost instantly in short, conversational form. Or there might be a pause and a longer, more structured answer. That difference is not just tone. It reflects how much computational effort the system has decided to spend.

The change isn't random; it helps Open AI solve a real problem. Though powerful, the most advanced AI models are also slower and more expensive to run. Using them for every single request would make your experience of using Chat GPT sluggish and costly.

Open AI can deliver something that feels both fast and capable by mixing up the models.

On the other hand, it makes for less predictable results. Two people using Chat GPT at the same time may not be using the same underlying system, even if their screens look identical.

One person might be routed to a lighter model, another to a heavier one, based on subtle differences in their prompts or usage patterns. The result is an experience that can vary in ways that are hard to explain from the outside.

Open AI introduces new Chat GPT 5.3 Instant to ‘reduce the cringe’ for users

Chat GPT's hidden creativity levers and how to flip them

Open AI takes steps to reduce ‘teaser-style phrasing’ in Chat GPT responses

You might not even know about all of the models available within Chat GPT anymore. They've been hidden in the configure menu, tucked away in settings that most people never open.

That's where the legacy models can still be accessed and where automatic switching can be turned off. You can even adjust how much effort the system applies when reasoning.

For casual users, this probably does not matter much. The experience feels smoother. You type something, you get an answer, and it generally works. The system takes care of the details.

For more attentive users, the change might feel slightly disorienting. If Chat GPT suddenly seems less detailed or more hesitant, it might be that it switched to a different model without telling you. In some cases, usage limits can trigger this kind of shift, quietly downgrading the level of reasoning applied to your prompts.

That creates a small but noticeable gap between expectation and reality. Many people assume they are interacting with a single, consistent intelligence. In truth, they are interacting with a flexible system that constantly adjusts itself.

This approach works because it removes friction. Most people do not want to manage settings or learn the nuances of different models. They want results. But when you no longer choose the model directly, you also give up some influence over how your answer is generated. That's not relevant for everyday questions, but it can matter when precision or depth becomes important.

If you want to ensure the model you're using then click on the model selector menu then look in the Configure setting. Inside there you'll see a Model selector.

The next phase of AI may not be about choosing the smartest model. It may be about understanding when and why the system is choosing for you. And if you want to take back control, the answer is sitting in a settings menu that most people never open.

Follow Tech Radar on Google News and add us as a preferred source to get our expert news, reviews, and opinion in your feeds. Make sure to click the Follow button!

And of course you can also follow Tech Radar on Tik Tok for news, reviews, unboxings in video form, and get regular updates from us on Whats App too.

➡️ Read our full guide to the best business laptops

1. Best overall: Dell Precision 5690
2. Best on a budget: Acer Aspire 5
3. Best Mac Book: Apple Mac Book Pro 14-inch (M4)

Eric Hal Schwartz is a freelance writer for Tech Radar with more than 15 years of experience covering the intersection of the world and technology. For the last five years, he served as head writer for Voicebot.ai and was on the leading edge of reporting on generative AI and large language models. He's since become an expert on the products of generative AI models, such as Open AI’s Chat GPT, Anthropic’s Claude, Google Gemini, and every other synthetic media tool. His experience runs the gamut of media, including print, digital, broadcast, and live events. Now, he's continuing to tell the stories people want and need to hear about the rapidly evolving AI space and its impact on their lives. Eric is based in New York City.

You must confirm your public display name before commenting

1 Chat GPT might not be using the model you think — and it’s also hiding others in settings

2 These are the top 6 4K Blu-rays I recommend from 2026 so far — and I've tested them all

3'I pretended like I wasn't afraid, but I was' — Ready or Not 2: Here I Come star breaks down that unhinged ending and one killer twist you definitely missed

4‘The bleeding edge of computing power for everyday creatives’ — this is why I’m upgrading to the Mac Book Air M5 after years of making do with my elderly Pro

5 Lenovo Ryzen 7 laptop with 16GB RAM and 512GB SSD is just $430 at Best Buy — but it won't be this price for long

Tech Radar is part of Future US Inc, an international media group and leading digital publisher. Visit our corporate site.

© Future US, Inc. Full 7th Floor, 130 West 42nd Street, New York, NY 10036.

Key Takeaways

• Chat GPT might not be using the model you think — and it’s also hiding others in settings

• There's a whole rotating cast of models working in the background

• When you purchase through links on our site, we may earn an affiliate commission

• Chat GPT now uses a simplified model picker that can automatically switch between different AI models

• Users may not realize that different prompts can trigger different underlying models