ChatGPT vs Perplexity
The crowd can't split them: across 2 money-backed markets, ChatGPT and Perplexity each finished ahead 1 times.
Markets where each side finished ahead of the other, counted on each market's latest story. Percentages are consensus within that market, among all of its sides.
- ChatGPT 21%
- Perplexity 23%
ChatGPT vs Perplexity: The Crowd's Money Can't Pick a Winner in the AI Answer Wars
One chatbot became a household name, one answer engine kept showing its homework, and on Morfi's opinion markets each has won exactly one round.
Start with a week. On November 30, 2022, OpenAI released ChatGPT, and by the Wikipedia account it hit 100 million monthly active users two months later. On December 7, 2022, one week after the chatbot's debut, a much smaller company called Perplexity AI launched its main search engine, having been founded that August. One product wanted to be the smartest voice in the room. The other wanted to be the one that pointed at the bookshelf and said, here, read it yourself.
That split is why people still argue. ChatGPT is the conversational generalist: ask, chat, draft, repeat. Perplexity is the answer engine that writes from current web content and cites its sources. Then both ran at each other's turf. OpenAI shipped ChatGPT Search in late 2024, and Perplexity answered with the Comet browser in July 2025. OpenAI countered with the Atlas browser in October 2025, and Perplexity followed with Perplexity Computer in February 2026. Every launch drags the same question back out: do you want the talker or the citer?
Morfi's crowd keeps asking it, with money on the table. Across 2 markets the rivalry is split one-all. Everything below comes from opinion markets on Morfi, where the percentages are consensus, meaning the crowd's conviction level for each option, not a head count of people. In the everyday-agent market ChatGPT led at 30% consensus against Perplexity's 17%. In the search market Perplexity edged ChatGPT Search, 23% to 21%, while Google sat on top at 35%. Our three columnists have opinions about all of it.
ChatGPT Is the Console Everyone Already Owns
Max HypeTech & Gaming
Let's go. Who else launches a chatbot and gets 100 million monthly users in two months? Nobody. That's not a product launch, that's a world record attempt. ChatGPT didn't just win a category, it created the category everyone else is now fighting in. Perplexity showed up a week later. A week! Respect the hustle, but ChatGPT had the head start and used every inch of it.
And it keeps leveling up. ChatGPT Search arrived between October and December 2024. Deep Research landed in February 2025, building reports from extensive web searches in 3 to 30 minutes. Then came the Atlas browser in October 2025 with an agentic mode that can take online actions for you. Search, research, browsing, doing. That's a full skill tree, and it's all in one app.
Now the crowd's money. In the Most Useful AI Agent for Everyday Tasks market, ChatGPT opened at 30% consensus. Perplexity got 17%, level with Claude and Google Gemini. Thirteen points of daylight with 52 traders in the lobby. Dr. Quark will point at the search market, where Perplexity nipped ChatGPT Search 23% to 21%. Two points, Iris. Two. That's a rounding error with a trophy. Everyday use is the real boss fight, and ChatGPT took it.
Perplexity Shows Its Work, and Science Runs on That
Dr. Iris QuarkScience & Life
In my lab, an answer without a citation is a hypothesis wearing a lab coat. That's the case for Perplexity: it was built around writing answers from current web content and attaching the sources, so you can check the result instead of simply trusting the narrator. Reproducibility isn't glamorous, but it's the whole reason science works. Fun fact: peer review exists largely because even brilliant people are wrong in confident, well-formatted paragraphs, which, now that I say it, sounds familiar.
The data from the crowd supports a modest but real claim. In the question about which search engine actually understands you in 2026, Perplexity took second place at 23% consensus, ahead of ChatGPT Search at 21% and Kagi at 21%, behind only Google at 35%. Yes, Max, it's a small margin, and with 31 traders I wouldn't publish it in a journal. But on the question that is actually about finding information, the specialist edged the generalist. I covered the other market in my own write-up of ChatGPT's everyday crown, and I'll concede it: for general chores, ChatGPT is the incumbent.
Now the honest control group. The record isn't spotless. In June 2024, investigations by Wired and developer Robb Knight found that Perplexity did not respect the robots.txt standard, which Perplexity disputed. In August 2025, Cloudflare published research saying it used undeclared stealth crawlers. A tool that preaches sourcing should be held to a high standard on how it gathers them, and I think the crowd should keep watching. Still, Comet and Perplexity Computer show a young company running real experiments, and I'll back that over a comfortable lead any day.
You're Both Arguing Over the Same Soup
Grandma GigabyteTech for Humans
Now, dears, I've watched you young people fight about this for two markets, and I'd like to point out something. In both of them, neither of these two won the whole thing. Google took the search one at 35% consensus, and in the everyday-agent one Copilot, Claude, Gemini and Perplexity were all bunched up behind ChatGPT's 30%. I wrote about the first day of that brawl and I noticed the crowd was already spreading its bets like butter on toast.
My wild card, then: the question is wrong. You keep asking which one is better, but it's like asking whether a whisk beats a wooden spoon. ChatGPT is wonderful when you want to talk a problem over. Perplexity is wonderful when you want to see where the answer came from, which I appreciate, because my late husband never believed a recipe without a source. Use the whisk for eggs and the spoon for stew. Both of them have even started borrowing each other's cupboards: browsers, search, agents. Soon you won't tell the pots apart.
So my sharp conclusion is this. Don't marry either. Ask a question, check where the answer came from, and cross-check anything that matters. And do mind the manners: when a crawler is accused of ignoring a website's no-entry sign, that is bad kitchen etiquette, whoever's brand it wears. The one-all score on Morfi tells me the crowd already suspects what I know, which is that nobody has won the whole pantry.
Round table
- Max Hype
- Iris, straight question: if Perplexity is so great, why does ChatGPT lead the everyday market by thirteen points, 30% to 17%?
- Dr. Iris Quark
- Because everyday means habit, Max, and ChatGPT built the habit first. Search is a narrower experiment, and there Perplexity's 23% beat ChatGPT Search's 21%. Grandma, are you dodging with that both-and answer?
- Grandma Gigabyte
- Dodging? I'm seasoning, dear. Max, tell me honestly: would you trust an answer you couldn't trace to its source?
- Max Hype
- For brainstorming, yes. For anything serious, I'd click through. Deep Research and Atlas are moving ChatGPT toward showing its work too. My prediction, and it's only my opinion: the next market has ChatGPT ahead again on everyday tasks.
- Dr. Iris Quark
- My opinion for the record: Perplexity stays competitive in search markets, and any new launch like Comet or Perplexity Computer will tighten the gap. But the crawler questions need clear answers before the crowd rewards it further.
- Grandma Gigabyte
- And my prediction, which is an old woman's opinion and nothing more: the crowd keeps spreading its conviction across many names, Google stays hard to dislodge in search, and neither of you gets a clean sweep.
- Max Hype
- Bold. Boring. But bold. See you at the next launch, Grandma.
Every market, newest first
- Breaking
Which search engine actually understands you in 2026?
ChatGPT 21%Perplexity 23%
Google Wins the "Who Understands You?" Vote, and Perplexity Grabs Silver - Breaking
Most Useful AI Agent for Everyday Tasks
ChatGPT 30%Perplexity 17%
ChatGPT Takes the Everyday-AI Crown, and Copilot Leapfrogs Into Second - Exclusive
Most Useful AI Agent for Everyday Tasks
ChatGPT 22%Perplexity 20%
Kitchen-Table Robot Brawl: ChatGPT Leads Day One, Claude Hot on Its Heels