I tested the new ChatGPT-5.4 mini and nano models and I didnt expect them to be this powerful

Date:
Wed, 18 Mar 2026 13:49:12 +0000

Description:
OpenAIs new GPT-5.4 mini and nano models bring a faster, more thoughtful ChatGPT experience to free users than I expected.

FULL STORY ======================================================================Copy link Facebook X Whatsapp Reddit Pinterest Flipboard Threads Email Share this article 0 Join the conversation Follow us Add us as a preferred source on Google Newsletter Tech Radar Get the TechRadar Newsletter Sign up for
breaking news, reviews, opinion, top tech deals, and more. Contact me with news and offers from other Future brands Receive email from us on behalf of our trusted partners or sponsors By submitting your information you agree to the Terms & Conditions and Privacy Policy and are aged 16 or over. You are
now subscribed Your newsletter sign-up was successful An account already exists for this email address, please log in. Subscribe to our newsletter Youve probably read about Sam Altman extolling the virtues of ChatGPT-5.4 Thinking as his favorite model to talk to. Thats great if youre a Plus or Pro user, because you get access to that model. The rest of us on Free or Go accounts only have access to a version of ChatGPT based on GPT-5.3.

Until today, that is. Today, OpenAI introduced GPT-5.4 mini and nano, describing them as our most capable small models yet, and the good news is that theyre available to Free and Go users in ChatGPT via a new Thinking option in the Plus (+) menu. Article continues below You may like OpenAI introduces ChatGPT 5.4 Thinking for solving bigger problems Switching ChatGPT to voice makes for a much better experience on mobile OpenAI introduces new ChatGPT 5.3 Instant to reduce the cringe for users Let me think about that
You cant choose GPT-5.4 mini or nano directly in the model selector in ChatGPT. Instead, you select the Thinking tool, and ChatGPT starts using them behind the scenes.

Even so, mini and nano bring a bit of GPT-5.4s thinking magic to free users, and Ive got to say, Im impressed with what Ive seen so far.

Thinking models work best when the answer isnt obvious, requires multiple steps, or benefits from careful reasoning rather than quick recall. To test the new feature, I asked my free ChatGPT account to: Plan a 5-day trip across Europe with a 800 budget, minimizing travel time and including 3 specific cities.

I wont paste the full travel plan here because it was quite long, but while the standard ChatGPT response was perfectly good, it was more of a bullet-point list. Get daily insight, inspiration and deals in your inbox
Sign up for breaking news, reviews, opinion, top tech deals, and more.
Contact me with news and offers from other Future brands Receive email from
us on behalf of our trusted partners or sponsors By submitting your information you agree to the Terms & Conditions and Privacy Policy and are aged 16 or over.

By contrast, when I selected the Thinking tool, ChatGPTs answer contained far more detail and gave me more insight into why it had chosen certain routes
and places to visit. It left me feeling more confident that this was a genuinely well-thought-out plan.

In short, the answer felt more human. Testing the new models To test the models with a prompt involving multiple dimensions, I used: I have 10,000, no coding experience, and 6 months. Design a realistic plan to generate income online, including risks, timelines, and fallback strategies. What to read
next ChatGPTs new settings range from corporate calm to chaotic bestie I tested ChatGPT Translate against Google Translate here's my winner The 5 prompts that reveal lingering 'cringe' in ChatGPT 5.3

Again, the two plans it produced werent wildly different, but the execution was miles apart. Standard ChatGPT gave me a brief outline broken down into bullet points, while the Thinking version fleshed everything out and
explained the reasoning behind its suggestions.

Ill give you a plan that balances speed, realism, and risk control, not hype, it assured me. It also gave me two tracks to run in parallel Track A for fastest income and Track B for building a scalable asset.

Both versions of ChatGPT included fallback strategies and risk analysis, but the Thinking version felt more solid and reliable. (Image credit: Getty Images/VCG) Open and reliable When OpenAI announced GPT-5.4 mini and nano, it was easy to assume theyd be compromises lighter, cheaper versions of the
real model designed mainly for background tasks and developers.

After actually using them, Im not so sure.

After testing the new models across everyday prompts writing, reasoning, and quick problem-solving what stood out wasnt what they couldnt do. It was how often they felt almost indistinguishable from the full ChatGPT-5.4 Thinking model, just noticeably faster.

When I asked the full-fat ChatGPT-5.4 Thinking model on my Plus plan the same questions, it did a lot more of exactly that thinking. Very slowly. Yes, it gave me more feedback on the decisions it was making, but it was essentially the same plan that GPT-5.4 mini or nano came up with. It just took longer to get there.

To be fair, the result was more personal and more detailed when I used a ChatGPT Plus account. But the fact that some measure of that thinking power
is now available on a free ChatGPT account is seriously impressive.

I hope Plus users get access to them soon, too, because the speed boost feels well worth the slight trade-off in depth. Follow TechRadar 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 TechRadar on TikTok for news, reviews, unboxings in video form, and get regular updates from us on WhatsApp too. The best business laptops for all budgets Our top picks, based on real-world testing and comparisons

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 MacBook: Apple MacBook Pro 14-inch (M4)



======================================================================
Link to news story: https://www.techradar.com/ai-platforms-assistants/chatgpt/i-tested-the-new-cha tgpt-5-4-mini-and-nano-models-and-i-didnt-expect-them-to-be-this-powerful


--- Mystic BBS v1.12 A49 (Linux/64)
* Origin: tqwNet Technology News (1337:1/100)