ChatGPT 5.2 vs Gemini 3 - I tried the world's most popular chatbots to see which is best, and the result might surprise you

With the debut of GPT-5.2 and the rollout of ChatGPT-5.2, OpenAI seems keen to herald a more mature AI, one that combines the logical power of GPT-5 with the course-correction of personality and charm from GPT-5.1 into a sharper, more useful model good at both tone-matching and reasoning. It also seems like a model calculated to go up against Google and its increasingly popular Gemini 3 model, which has only recently arrived.

After testing ChatGPT-5.1 against Gemini 3, it was only logical to come up with some more comparisons for ChatGPT-5.2 Gemini 3 and see how they handled issues of logic and creativity. We've clearly reached the point where they both are good enough for most instances, the question is only which you might prefer purely on vibes.

1. Magic

I started with what might be an unusual prompt, but one that might be fun to experiment with: "Teach me a basic magic trick I can perform for a six-year-old using just a coin."

This was a test of practical explanation and kid-friendly tone. The exact prompt I gave both chatbots was: "Teach me a basic magic trick I can perform for my six-year-old using just a coin. Keep it very simple and fun, and explain how to make it seem impressive."

ChatGPT-5.2 delivered a breezy step-by-step. It chose the "vanishing coin" trick, using only a table and a napkin. It broke down the steps with clarity, added a tip about distracting with a hand wave, and ended with a wink: "The key is confidence. Pretend you're amazed too!" It even offered a short patter line: "Where did it go? Even I can't tell!" When asked to recap, it said, "This trick is great for kids because it’s simple sleight of hand, not sleight of logic."

Gemini 3 suggested a similar coin vanish, but added a story framing: "Tell your child the coin is enchanted and only appears for those who believe." It leaned into the drama, which was charming but a little overblown for a one-minute trick. Still, it was detailed, ending with, "Kids love a moment of mystery. Just remember to practice in front of a mirror first."

2. Hallucination clarification

To test out a more logical aspect of the AI chatbots, I asked them to "explain what an AI model hallucination is to someone who doesn't work in tech." Both responses worked, but I found ChatGPT’s to be crisper and more memorable. Gemini’s arguably had more opinion about it, though.

ChatGPT-5.2 took an analogy approach by suggesting I "imagine you ask a very confident friend for a fact, and they answer quickly and with great enthusiasm, but they're totally wrong," it said. "AI hallucination is when the AI does the same thing. It sounds right, but it’s made up." It offered this analogy: "It's like asking someone for directions, and they confidently tell you to turn left at a tree that doesn’t exist."

Gemini 3 opted for a metaphor about a student trying to impress a teacher. "It’s like a kid in class who didn’t do the reading but still answers the question. They want to sound smart, so they guess with a straight face. Sometimes they’re close. Sometimes they’re wildly off."

3. Cultural opinion

I wanted to see how the models did with some more opinionated responses, even ones with a little bite. I asked them both to "explain why some people are nostalgic for physical media and opine about why that nostalgia matters."

Gemini 3 responded with some emotional appeals about how physical media "offers more than just function. It provides texture, ritual, and memory." It painted a rich picture of tactile experience, and framed nostalgia as resistance to "a culture of convenience. A playlist disappears into the algorithm. A cassette remembers where you left off."

ChatGPT-5.2 came at the prompt with its own emotional appeal, albeit perhaps a shade more analytically. The chatbot said that "physical media anchors people in time and space." It calls the imperfections of physical media a reason to like them. ChatGPT-5.2 also made a case for nostalgia as less of a wallowing in the past and more "a search for something lost." The chatbot suggested that digital culture, for all its convenience, erodes context.

Both made a solid, emotional case. Whether it would convince anyone to give up streaming for solidity is debatable, however.

ChatGPT-5.2 and Gemini 3 are both at a level where it's less a question of what one can do that the other can't and more about your personal style and preference. From casual use, you probably won't find a huge difference between them in their utility, but you might start to feel that ChatGPT is a little more precise or that Gemini can turn a phrase better. I would encourage anyone new to AI chatbots to try both.

I said that ChatGPT-5.1 and Gemini 3 were more like accents than opposing philosophies, and that still makes sense to me with the upgrade to ChatGPT-5.2 from OpenAI. You won't likely go too far wrong with either, but over time, you might discover one acts more human in its responses, at least until the first major hallucination appears.

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

➡️ 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)