Everyone forgot about Perplexity, but it still does this one thing better

Not long ago, Perplexity was the AI tool everyone was talking about. It was labeled a Google killer, and everyone thought it was poised to change web search forever. Then ChatGPT added web browsing, Google launched Gemini with real-time search, and suddenly, the hype started to die down.

Perplexity itself contributed to that. If you bought an annual Perplexity subscription, you were lied to, and the chatbot had started giving wrong answers on purpose. But despite all of that, it still does one thing better than most AI assistants or chatbots you’ll find on the market. Writing Perplexity off entirely misses what it still does remarkably well—and that’s one of the biggest reasons to continue using it.

Everyone ran to the shiny new AI toys

How Perplexity quietly disappeared from the conversation

The reason Perplexity fell out of the spotlight is pretty understandable and obvious. OpenAI built web search directly into ChatGPT, and because most people were already using ChatGPT for everything else, switching back to a dedicated tool added friction no one wanted. Gemini, on the other hand, had Google’s entire search index baked in. Grok had real-time access to X. The unique selling proposition that Perplexity had spent years building—the idea that an AI search engine with sourced answers was genuinely novel—stopped feeling unique. The competition had caught up, and to some extent, surpassed Perplexity.

Research is Perplexity’s main strength. For tasks like creative writing, coding, and long-form content generation, Perplexity just isn’t the right tool. To make matters worse, it started aggressively rate-limiting Deep Research—the one feature almost everyone used Perplexity for, especially paid users who were promised entirely different limits when they signed up. Local LLMs with internet access also started replacing Perplexity—making that extra $20 per month hard to justify.

OS

Any

Developer

Perplexity

Price model

Free, $20 per month Pro subscription

Initial release

December 7, 2022

Perplexity is an AI-powered search engine that combines web search with conversational answers. It scans reliable online sources in real time, summarizes key information, and provides cited responses. Designed for clarity and accuracy, Perplexity helps users quickly find, verify, and understand complex topics through concise, trustworthy explanations.

But it still beats them at this

Fast, sourced answers without the usual fluff

Access to the internet might not be unique to Perplexity anymore, but one thing is: it treats source transparency as a first-class feature, not an afterthought.

Every answer Perplexity gives you is wrapped in numbered inline citations that are directly linked to the original source. This isn’t a vague list of links at the bottom and not a footnote section you have to scroll to. The citation is right next to the sentence, tied to the specific claim it supports, so you can click it and immediately verify whether what you just read is actually true or the AI is just hallucinating. It sounds like a small thing, but it completely changes how you interact with the answer and where you can use the tool.

Sure, ChatGPT and Gemini have gotten better at including sources, but they still tend to deliver information with a confident, essay-like tone that discourages you from questioning them. I love using Claude, but it makes some painful mistakes that can’t be ignored when it comes to output accuracy. The winner here is clearly Perplexity, especially when you want quickly verifiable answers and don’t trust the AI with your eyes closed.

There’s a reason people joke about AI hallucination, and that’s primarily because these tools tend to give answers with the same authority, whether they’re quoting a peer-reviewed study or making something up. LLMs tend to hallucinate more in certain tasks, so having the source cited in the same sentence as the claim massively helps in verification.

That advantage matters more than you think

Why accuracy changes how you actually use AI

This isn’t just a nerdy detail about UI design. It fundamentally changes what you can use AI search for. If you’re a journalist fact-checking a claim, a student trying not to cite something that doesn’t exist, or just someone who doesn’t want to be confidently misled about something you’re researching, the ability to trace everything back to its source is paramount. It builds a relationship between you and the information that some more conversational AI tools don’t support as much.

Perplexity claimed to process 780 million search queries in May 2025, and that was before it launched its AI-powered browser Comet. That may not be as high a figure as seen by its competitors, but it’s certainly not a sign of a tool falling off the edge. These are the numbers of a tool that a specific kind of user—the kind who actually cares where information comes from—keeps coming back to because nothing else has genuinely replicated what it does.

Its best feature is how unexciting it is

A quiet strength that makes it genuinely useful

The irony is that the feature that made Perplexity famous in the first place is also the feature that gets the least attention among the endless “which AI is better” debates you’ll find online. Reviewers love comparing report length and writing style, users like comparing rate limits and subscription prices, and then there are comparisons with the competition.

You can debate whether ChatGPT writes better prose, whether Gemini has a bigger index, or whether Claude generates more robust code. Those arguments are largely a matter of use case. But when it comes to giving you an answer and them showing its work, cleanly and immediately, without making you dig for the source, Perplexity is still the benchmark, and I don’t see that changing anytime soon.