There’s something wild happening in the world of AI search—Perplexity, a relatively fresh face in the AI space, just crossed the $100 million mark in annual revenue. No fluff, no inflated ARR tricks—real revenue from real users. And what’s even more impressive? They did it in just 20 months.

If you’re still thinking Perplexity is just “another AI chatbot,” it’s time to look again.

CEO Aravind Srinivas is not playing small. From day one, the team at Perplexity wasn’t trying to be the next Google. They were trying to replace the need for Googling altogether.

It started with a search engine that actually gives you answers, not just links. Now, they’re moving toward what Srinivas calls “agentic search.” Think less “What’s the best hotel in Bangkok?” and more “Book me a boutique stay in Sukhumvit next Friday with breakfast included.”

That’s not a hypothetical. You can now book hotels directly inside Perplexity. They’ve partnered with TripAdvisor and Selfbook to make it happen. Discounts and seamless transactions? Already part of the plan.

This isn’t about becoming Expedia. It’s about building a smart assistant that just happens to know everything and can actually act on it.

Let’s talk about the next big leap: Comet.

Comet is Perplexity’s upcoming agentic browser, built on Chromium. Details are still fuzzy, but one thing’s clear—it’s not your regular browser. For one, it might come with YouTube ads disabled by default (yes, you read that right). But more importantly, it’s built to give AI agents deep access to your browser.

Why? Because AI agents can’t be useful if they’re stuck on the sidelines. To actually help you shop, plan, organize, and execute tasks on iOS or any locked-down platform, they need control. Comet is that bridge.

Aravind’s long-term vision? An AI-powered personal assistant that can do the work of an executive assistant. Not just summarizing your emails, but acting on them.

And if you can’t afford a real-life EA (which, let’s be honest, most people can’t), this could change your daily life. We’re talking grocery orders, travel plans, finding gifts, managing subscriptions—all handled with a sentence or two.

Perplexity’s valuation hit $9 billion a few months ago. Now it’s reportedly shooting for $18 billion in a fresh funding round.

Let that sink in.

A year ago, it was valued at $1 billion. That’s 18x growth in 12 months. And this isn’t meme-coin hype—it’s grounded in actual traction, usage, and revenue.

Still, some people are skeptical. After all, the big boys—OpenAI, Google, Meta—are building search into their own AI systems. What happens when ChatGPT or Gemini does everything Perplexity does?

Srinivas isn’t blind to that. In fact, he’s betting on the opposite.

Instead of trying to be everything, Perplexity is going deep into specific use cases. That new “answer mode” they launched? It can fetch hyper-specific results—like a particular jacket or a remote job listing—without making you scroll for hours.

This kind of vertical focus could be what gives them an edge in the long term. It’s not about having the most features—it’s about building the most useful ones for the people who actually need them.

Now here’s the twist: Google is still technically ahead in almost every metric—from technical model strength to user base to market dominance. And yet, Google Search feels increasingly… clunky. Corporate. Ad-heavy. Stale.

Compare that to Perplexity, which just launched a blazing-fast answer mode, letting users find hotels, clothes, jobs, and videos directly through the interface. No clickbait. No ads. No 5-minute scrolls.

The secret? They’re not trying to copy Google. They’re building something completely new

.

Let’s be honest: Google DeepMind built some of the most powerful models in the world. But what happened next?

OpenAI launched a doodle feature, and that stole all the headlines. Perplexity shipped a hotel booking integration, and suddenly it felt more useful than Google Search.

Even investors are noticing this gap. One quipped:

“You build the smartest model ever made… and users are still stuck on the ‘Inspire Me’ tab on Gemini.”

Here’s the deeper problem: Google isn’t failing because it’s behind on tech. It’s failing because it’s forgotten how to build for real people.

• Gemini Advanced is locked behind a paywall—while ChatGPT and Perplexity give value upfront.
• Templates and UI on Gemini feel dry, uninspired.
• Core features like NotebookLM are buried in experimental tools, far from mass users.

Meanwhile, Perplexity feels sharp. Fast. Personalized. Every update feels like it’s been tested with actual human beings.

Perplexity’s approach to monetization is also quietly brilliant. Instead of drowning the interface in ads like Google, they’re integrating native transactions—you can book hotels, buy products, or even access partner discounts directly within the platform.

This vertical integration model (think TripAdvisor meets ChatGPT) might just become the future of search engine monetization.

It’s clean. It’s user-first. And it positions Perplexity as a trustworthy platform—not an ad machine.

Perplexity isn’t the only company proving that AI can generate serious revenue without mass adoption. Another fascinating case is a health-tech AI tool built specifically for nutritionists. It serves just 3,000 customers and still made $50 million in annual revenue.

How? Because it solved an actual problem.

This is where AI is headed: real tools for real people, not just demo-stage hype. Whether it’s clinicians using Freed AI to automate note-taking or nutritionists streamlining their meal planning, vertical AI is quietly becoming the most lucrative segment in the industry.

Crossing $100M in revenue this early isn’t just a milestone for Perplexity—it’s a sign that users are ready for something new. A tool that doesn’t just “search” the internet, but acts as your assistant on it.

Google still has the throne. But with every seamless experience, every natively booked trip, every AI-powered answer that actually solves something—Perplexity is quietly building a kingdom of its own.

And if Google doesn’t wake up to the user experience gap soon, it may end up being the smartest AI company… that no one wants to use.