Curious Companion: An Introduction to Perplexity, the Answer Engine

LISTEN TO THE EPISODE

In this episode I get curious about Perplexity, a name I’d heard used a ton in the AI space but never actually knew what the heck it was. I break down what it is, how its citation-first approach set it apart from ChatGPT early on, and what it can actually do today: Deep Research, Labs, and its push into the agent space with Computer, Brain, and Comet. I also cover the legal issues that are serving as a sneak peek into the future of agent wars, and give my honest take on whether Perplexity is worth using.

I Never Knew

Today we’re diving into Perplexity, a name I’ve heard mentioned in the AI space for quite some time now, but never knew what the heck it was, so I decided to get curious and do a little learning.

Based on what I found, this is definitely NOT an “I’m talking about this because I recommend this tool” kind of episode. I’ve yet to actually use Perplexity outside of the teeny bit of tinkering I did so that I could create this episode, and I honestly don’t have any use cases for it off the top of my head. I’m also not convinced that it’s any better than Claude for my specific needs, but TBD.

Finally, I’m also not sure if Perplexity is on its way out the door (lots of negative talk on Reddit, and we know what happens when companies stray from their original service), but we’re curious people here, so let’s dig in.

So What Is Perplexity?

Per its own website: Perplexity is an AI-powered answer engine that provides accurate, trusted, and real-time answers to any question in the form of conversational summaries backed by real-time web citations. The translation and main point I want to highlight: it provides citations (links) with its answers.

Worth highlighting up front for my techies in the audience, Perplexity isn’t its own foundation model. It’s a harness that orchestrates other companies’ models (GPT, Claude, Gemini, Grok, plus some in-house fine-tuned ones) on top of live web search. The user selects which model they want to use and then inputs their prompt as usual. Of note, only users on a paid plan can actually select the model they use; free tier users must use whatever Perplexity has set as the default model.

Perplexity offers two paid tiers in addition to its free tier:

Perplexity Pro – $20/month (if billed monthly)
Perplexity Max – $200/month (if billed monthly)

Perplexity seems best for researchers, consultants, educators, analysts, and anyone whose deliverable has to survive someone asking “where’d you get information from.” It’s absolutely far less useful for creative writing, heavy coding, or huge single-document deep dives.

Perplexity has been around for a minute, and actually went live on December 7, 2022 (7 days after ChatGPT launched), differentiating itself from the competition with those citations that I mentioned earlier. ChatGPT was busy making stuff up, and Perplexity was establishing itself in the market as a trustworthy resource.

However, that same “grab it from the web and reformat it with citations” approach has landed Perplexity in active lawsuits from the BBC, The New York Times, Reddit, and multiple Japanese newspaper companies over alleged copyright infringement.

So What Can Perplexity Actually Do?

Given that this is an introduction to Perplexity, not a deep dive, I’m gonna go over the big rocks regarding features. if you’re really interested and want the minutiae, as per always, go try it out for yourself!

Regarding basic functionality, you ask Perplexity something like you would any other LLM, and by default it runs live searches and gives you a synthesized answer with clickable sources. If you just said to yourself, “So does ChatGPT and Claude,” you would be 100% correct, and we’re going to talk about what that’s meant for Perplexity’s business model in a second.

For those wanting more answers with more substance, there is a Deep Research mode that will generate a source-backed report from the queried web material. Important note: this feature is only available to folks on a paid plan.

Perplexity can also create things like mini apps, dashboards, files, or other project-like outputs rather than just a written research report. This is known as a Labs task. Again, yes, Claude can do this, and that competition is what I want to talk about next.

Entering its Agent Era

Earlier this year, February to be specific, Perplexity ventured into the agent space and launched “Computer.” Remember: agents are AI programs that execute multistep tasks (they actually do things, not just answer questions) autonomously. For the deep dive on agents, check out episode 52.

The best way I can explain Computer is that it’s analogous to Claude Cowork. Computer is an agent that functions as an autonomous digital worker, and uses various frontier AI models (GPT, Claude, Gemini, etc) to execute multi-step workflows. As of July 2026 it’s expanding into Microsoft 365: Word, Excel, PowerPoint, Outlook, and Teams. Again worth noting, access is gated to a paid plan.

Two other names you’ll come across if you venture into the Perplexity ecosystem are Brain and Comet.

Brain: launched June 2026, is a memory system for Computer. What makes it different is that it remembers the agent’s own work (what succeeded, what failed), not your personal preferences. Perplexity’s own early numbers claim accuracy gains, but those are first-party, unverified figures, so take that statement with a grain of salt.
Comet is Perplexity’s own web browser, built with AI baked in. It can summarize pages, answer questions about what you’re looking at, and carry out multi-step tasks directly inside the browser.

Fun tidbit (and a glimpse at the impending agent wars): Amazon is suing over Perplexity’s Comet browser acting as a shopping agent on users’ behalf. Comet logs into a user’s Amazon account with the user’s permission…but Amazon itself never authorized it. It sounds to me like Amazon is mad about lost ad revenue (if the Agent is shopping instead of the human, the human never sees the ad), and also wants folks using their own agent, Rufus, but what do I know.

Should You Bother With It?

As per always, play around with it and make the final call yourself. The free tier is a perfectly fine place to start, and clearly the best use case is questions where you want a source attached.

I don’t pay for Perplexity, and I haven’t used it outside of the little bit of poking around I did in order to create this episode. To that end, unlike with ChatGPT and Claude I can’t say if it’s worth it or not to pay for it and get those pay-gated features.

I think for the right user with the right use cases, Perplexity could be a helpful tool, but I personally don’t see myself adding it to my tech stack anytime soon (especially given Claude’s Research mode that I covered in episode 53.

But, again, we’re curious folks here, so if any part of you thinks it might be helpful, definitely give it a try yourself and see if it’s worth it for you.

How I Used AI This Week

Each episode I share a quick example of how I used AI that week.

This time I used Claude Code to build a custom content creation tracker for the members of my newest program, Crew 52.

I’ve written in the past about how I used Claude Code to create the custom dashboard I use to manage my messaging clients, and I used the exact same scaffolding for this. The difference this time is that the tracker is client-facing and needs to withstand the stress test of external users, not just me.

So far so good, and the build was much faster this time despite the increased complexity. I completed the entire thing over the course of two afternoons with basically no issues, and honestly the main reason it took two days instead of one was because I hit the usage limits twice and had to wait for it to reset (I’m still on the $20/month tier, and I used Sonnet 5 to build this).

For those of you looking to build anything, I cannot overstate the benefit of taking a good amount of time to plan things out with Claude, and then having Claude Code build in multiple phases instead of trying to one-shot it. Phases allow you to change things along the way without needing to do some massive overhaul of one part that inevitably winds up breaking another. (I still can’t believe that we have access to this kind of technology for $20/month.)

If you’re thinking about hopping on the vibe coding train, please know that repetition will be kind to you. Get to building and I promise it will get easier.

Da Wrap-up

Perplexity, put simply, is an answer engine, aka a tool that searches the live web in response to your question and returns a citation-backed answer. Those citations were the differentiators, and now they’re basically table stakes since ChatGPT and Claude do it too. It’s clear that Perplexity is attempting to stay relevant by doubling down on the agent space with Computer, Brain, and Comet, and we’ll have to see how it does with those lawsuits. Given how I personally use AI, I remain on the fence about its utility, but as per always your best bet is to try it out and make the call yourself.

As always, endlessly grateful for you and your curiosity.

Catch you next Thursday.

Maestro out.