Despite growing competition from AI chatbots, Google still dominates the search engine market. And looking at chatbots alone, ChatGPT is the clear winner. So how do you compete with that? Perplexity bills itself as a free AI-powered “answer engine,” and it’s picked up enough steam to score an integration on Samsung’s Galaxy S26 lineup. But how well does it do its job? To find out, I replaced my day-to-day Google searches with Perplexity AI queries for a week.
As an AI “research assistant,” Perplexity lets you to filter sources (“academic only,” for instance), while a Focus mode lets you narrow results to those drawn from a particular site or category. What stands out to me compared with rival AI chatbots is that Perplexity provides inline, numbered citations for every output. Also, unlike some of its competitors, Perplexity draws on real-time web results rather than relying primarily on training data, and it can leverage multiple models to answer queries.
However, to quote The Wire’s Omar Little, “You come at the king, you best not miss.” So did Perplexity hold up against the undisputed champ? Let’s dive in.
Setting Up My Parameters
To test its functionality, I put Google aside for a week and relied solely on Perplexity AI. As a tech and science writer, I use search a lot to provide context for topics or to deep-dive into topics like scientific principles and manufacturing histories. I’m also writing a science-fiction novel heavily based in real science and technology, set primarily in a massive swamp biome, so I use search a lot to dig into details about things like biodiversity in swamps (and military technology—the book has giant mechs, as well).
In my personal life, I use search pretty regularly to answer all the random questions that pop up in conversation throughout the day. I try to keep up with the news, so I often use search to gather additional context and fact check. I also read a lot of reviews—films, TV, games, products, and devices—and Google is my portal to that kind of criticism. I generally eschew AI for those purposes and rely almost exclusively on Google (and Ecosia, a similar search engine with an environmental mission), so switching to Perplexity exclusively was a dramatic (and illuminating) shift for me.
Google vs. Perplexity: What I’m Getting From Each
As the world’s most dominant search engine, Google uses a massive web index and advanced ranking algorithms to scour the internet for information. It responds to queries with brief AI summaries, knowledge panels, snippets, and a “People Also Ask” section with likely follow-up queries. Google search strives to provide relevant, personalized results by using your location and search history, and integrates seamlessly with the rest of the tech giant’s ecosystem, including Maps, Calendar, Drive, and YouTube.
As a combination of chatbot interface, research assistant, and search engine, Perplexity is a different beast entirely. It delivers highly focused results with minimal ads and clutter. Thanks to the AI’s integrations with real-time web and academic sources, each response includes citations with direct links to sources. And after each response, it also offers a selection of likely follow-up queries.
Note: I was using the basic version of Perplexity AI for this experiment. The Pro version adds the ability to choose which model to use for your queries, upload files, shop, and access other advanced features (though the basic version gives you access to a limited number of deeper Pro responses per day). A Perplexity Max subscription provides more unfettered access for AI power users.
(Credit: PCMag / Perplexity)
The Results
So what did I find after using Perplexity in place of Google search for a week? Predictably, each tool has its own strengths. However, the surprise to me was how much more I preferred Perplexity for the vast majority of use cases I tested.
Reviews and Criticism: Google
Google continues to shine for reviews and criticism. A big part of what I value about criticism is the subjectivity of a reviewer and the opinion delivered in their own words. To get that from Perplexity, I had to treat it like Google, forcing it to search for specific reviews and link me to them, which is clumsy in a chatbot interface. Because of AI’s specific idiosyncrasies, I also don’t trust it to summarize a reviewer’s taste, which is the context I like when reading a review. Instead, I again had to force the chatbot to link me to other work by the same reviewer for additional context, which is much easier via Google.
Similarly, searching for specific products is easier and more informative than searching in Perplexity, though in fairness, a lot of that kind of search I do directly on specific retailers’ website, like Amazon.
Get Our Best Stories!
Your Daily Dose of Our Top Tech News
By clicking Sign Me Up, you confirm you are 16+ and agree to our Terms of Use and Privacy
Policy.
Thanks for signing up!
Your subscription has been confirmed. Keep an eye on your inbox!
Summaries and Citations: Perplexity
In my testing, I found Perplexity was far superior at providing reliable, transparently sourced information about the topics I was researching, including principles of general relativity, microscopic robots, the scale of the universe and stellar phenomena, and how AI is being deployed to help fight fires.
Perplexity was excellent at handling complex data and condensing it into something more understandable to the layman. Its ability to summarize incredibly complex scientific studies was also impressive, and I found it very easy to modulate the language from super technical to incredibly simple, depending on my needs. Being able to source only academic resources was a huge boon, too, since my work requires me to use only highly reputable (sometimes peer-reviewed) sources.
Recommended by Our Editors
Citations were huge. Having clear, easily accessible sourcing for each and every data point made double-checking against hallucinations or digging deeper into a response really easy. I did run into an issue during one query where every citation was identical to the previous response, but the AI quickly corrected it when I pointed it out, and I never encountered it again.
Research: Perplexity
I also found it really helpful for personal research. Following the conflict in Iran has been difficult, but Perplexity was great for providing greater context on history, politics, and the factional makeup of many of the actors involved, as well as drawing from high-quality sources for more subjective analysis and forecasting.
This is where I found the follow-up refinement threads the most useful. Perplexity not only frequently predicted queries I would’ve followed up with myself, but actually suggested a number of follow-ups I wouldn’t have thought to ask, and they were really interesting and useful.
That said, it still comes with the typical AI caveats. I caught it hallucinating twice: it misreported the drone’s altitude ceiling while I was researching a story about recharging unmanned aerial vehicles in mid-air with lasers. At one point, it also cited a statistic from a 2023 study that did not exist. While the study was real, the specific data that was referenced was not.
Perplexity Is a Great Research Buddy (That You Need to Watch)
While I mostly used it for research and general web queries, it’s worth noting that Perplexity has a number of features tailored to specific users that I didn’t use, like market and finance tools. Another feature I didn’t get too deep into was its Spaces tool, which lets you organize searches and threads by topic or project, acting as a centralized hub for research and collaboration.
That said, I came away impressed with the features and functionality I tested. I can honestly say I’m a convert, and though I’ll continue to limit my overall AI usage due to environmental concerns, I will replace a significant chunk of my Googling with Perplexity searches. Google remains the map of the web, but for me, Perplexity is fast becoming its smart guide.
About Our Expert
Experience
Alan is an experienced culture and tech writer/editor with a background in newspaper reporting. His work has appeared in Rolling Stone, Paste Magazine, The Escapist, ESPN, PC Gamer, and a multitude of other outlets. He has over twenty years of experience as a journalist, author, and editor.
Read Full Bio
