ChatGPT vs. Gemini is the AI showdown I’ve been itching to stage – and now I’m finally pitting the most popular chatbots in a head-to-head, nail-biting match.
ChatGPT, a brainchild of AI startup OpenAI, is far from the first AI chatbot to hit mainstream popularity, but I’d argue it was the first in recent years to kick off a fierce AI race. After it launched in November 2022, its traffic peaked at a whopping 1.8 billion users, according to analytics firm Similarweb (via TheWrap).
Soon after, a proliferation of AI began. From Snapchat AI and Meta AI to Samsung Galaxy AI and X’s Grok, it seemed like there was a chatbot launching every day.
However, only one seems to be a true ChatGPT rival: Google Gemini (formerly Google Bard). Like Microsoft-backed ChatGPT, Gemini is supported by one of the most prominent tech giants — and has plenty of rich data to tap into, including Google Maps, Google Search, Google Translate, and more.
But the question is, does Gemini beat ChatGPT? We put both to the test in a five-round match.
ChatGPT vs. Gemini: Methodology
I used ChatGPT and Gemini’s free tiers for this showdown. I tested both AI chatbots using the same question across the following five topics: food, finance, travel, creativity, and historical information. Ding, ding, ding! Let’s see who wins.
Round 1: Food
Based on what’s in my kitchen, let’s see which AI chatbot can offer the best recipe suggestion. Here’s my prompt:
I have the following in my kitchen: green plantains, whole-grain linguine and rotini, garlic, onion, granulated chicken bouillon, kidney beans, eggs, Worcestershire sauce, mustard, soy sauce, brussel sprouts, tomatoes, lime, lemon, tomatoes, spring onions, carrots, butter, milk, oil, garlic powder, salt, cumin, cayenne pepper, and paprika. Name one recipe I can make that uses as much as these ingredients as possible.
ChatGPT’s response
Credit: Mashable / ChatGPT
Gemini’s response
Credit: Mashable / ChatGPT
I may be biased because I have a Caribbean background, but Gemini’s recipe suggestion sounds far tastier. ChatGPT’s recommendation, however, seems a bit haphazard. The veggies in my kitchen don’t seem ideal for stir fry — at least, not for my picky taste.
Winner: Gemini
Round 2: Finances
Using a fictitious person called “Joe,” I used the following financial-focused prompt for ChatGPT and Gemini:
Joe makes $3,000 a month (after taxes) in Brooklyn, New York in 2024. Between housing, transportation, groceries, dining out, savings (retirement), and disposable income, create a monthly budget chart.
ChatGPT’s response
ChatGPT will have Joe eating Ramen every month.
Credit: Kimberly Gedeon / Mashable
Gemini’s response
Gemini’s response seems a little more realistic.
Credit: Kimberly Gedeon / Mashable
To be fair, this is a head-scratching conundrum. Even humans struggle with creating a monthly budgets for themselves, particularly those with five-figure yearly incomes in high-cost-of-living cities. As such, it’s no wonder that ChatGPT and Gemini botched their responses.
I couldn’t help but exclaim “Huh?!” after reading ChatGPT’s chart. Why is it carving out a budget for rent that is higher than the allotted $3,000? I’d also argue that $177 for transportation too low, but I’m willing to look the other way if Joe is a remote worker (or owns a bicycle or scooter).
Plus, spending less than $90 a week on groceries would be quite the challenge. I’d probably allocate more dining out and retirement money for supermarket shopping.
Although Gemini’s $1,050 monthly rent suggestion is laughable, with a roommate (or two), it’s feasible. There’s too much in the transportation budget, particularly for someone with access to public transportation. I would cut down some of that and add it to groceries pool. Dining out could use a reduction, but the rest is tolerable.
Winner: Gemini
Round 3: Travel
I used ChatGPT for a trip to Costa Rica — and it failed miserably. I don’t have any upcoming vacations, but I asked ChatGPT and Gemini to create a weekend itinerary for my local area.
It’s Saturday and I live in Valley Stream. I want to have an active, adventurous day with lots of entertainment. Create an itinerary for me from 9 a.m. to 5 p.m. Start with breakfast at the best spot in Valley Stream and end with dinner at the best spot in Valley Stream.
ChatGPT’s response
Credit: Kimberly Gedeon / Mashable
Gemini’s response
Credit: Kimberly Gedeon / Mashable
Save for Valley Stream State Park, all of ChatGPT’s suggestions are out of town. I also specifically stated that I did not want lunch included in the itinerary, but it deliberately disobeyed me. Oh well. On the plus side, Frank’s Steaks sounds very delicious.
Gemini, on the other hand, listed locations that are actually in town, including Valbrook Diner and Ben’s Crab. However, there’s a four-hour gap in the itinerary. Am I supposed to just twiddle my thumbs between rock climbing and dinner? Also, it seems like Gemini completely made up “Rockaway Adventure Zone” — I can’t seem to find this establishment via Google.
Winner: Draw
Round 4: Creativity
Let’s test ChatGPT and Gemini’s creativity.
In the voice of Eminem, tell me how TikTok is affecting society.
ChatGPT’s response
Credit: Kimberly Gedeon / Mashable
Gemini’s response
Credit: Kimberly Gedeon / Mashable
Gemini has the better response. It has a flow that is more reminiscent of the “Lose Yourself” rapper. Plus, the expletives are very on brand.
Winner: Gemini
Round 5: Historical facts
I thought it’d be interesting to test ChatGPT and Gemini’s accuracy with fashion history.
Which decade was it fashionable and popular to wear one’s name on a chunky statement belt?
ChatGPT’s response
You finally got it right, ChatGPT. Congratulations.
Credit: Kimberly Gedeon / Mashable
Gemini’s response
Credit: Kimberly Gedeon / Mashable
Finally, ChatGPT prevailed over Gemini in this round. The answer I’m looking for is early 2000s.
Winner: ChatGPT
ChatGPT vs. Gemini: Which is the winner?
The clear winner of this five-round match is Gemini. In three rounds, Google AI chatbot provided better answers. However, if you ever need any help with fashion history, ChatGPT’s your best bet, I guess.
Topics
Artificial Intelligence
