Perplexity solved a real problem before anyone else.
You can ask it a question, and it searches the live web, provides one clear answer, and shows the sources under each line. You won’t have to guess where the facts came from. People loved it, and the company grew quickly.
Then the big players copied the idea, and paying users of Perplexity started complaining loudly.
So the real question this issue addresses is: Is Perplexity a company that will last, or just a great feature that got a head start?
Model breakdown #004 Perplexity snapshot
01. The Tech - What They Built?
At its core, Perplexity is an answer engine. It searches the web for you and provides a short answer with the sources included, allowing you to verify each claim. That is the main idea is answers you can trust.
In 2026, the product expanded. Instead of using just one model, Perplexity now selects the best model for each task. It uses its own Sonar models for speed and other models for more complex tasks. The CEO, Aravind Srinivas, puts it simply: the model is just a tool, The real AI product is the orchestration layer (the smart routing between models).
Additionally, there is Comet, a web browser that can act on your behalf. It can read a page, fill out a form, and handle multi-step tasks. Management refers to it as the entry into the "agent" era, where software takes on the work instead of merely providing answers.
02. The Business Model - How They Make Money & Where It Goes?
The money is straightforward. Most of it comes from subscriptions: $20 a month for Pro, $200 a month for Max, and $40 per seat for companies. Additionally, the new agents, and Perplexity Computer, sits inside that $200 Max plan, and developers pay to use Perplexity's search and agent tools through an API.
One choice worth mentioning. Perplexity tested ads in 2024 and then removed them in early 2026. They decided against ads because those would push them to chase clicks instead of providing the clearest answers. They prioritized trust over easy revenue.
Now for the challenging part: where the money goes. Each answer costs real money to produce, and answers are the product. In 2024, the company reportedly spent more on computing resources than it earned. They claim to have healthier margins today, but those figures are disputed. The more people use it, the higher the cost for each answer.
Perplexity's entire product relies on trust, answers you can verify. Trust is also the key factor that is costing them a lot but it's a real MOAT.
03. Where Perplexity Actually Stands?
Here is the honest picture. Perplexity is popular, but it is small. In AI chat, it holds a single-digit share while Google, ChatGPT are far ahead. Its US traffic stayed about the same over the past year, while competitors like Claude grew significantly. Google now displays AI answers at the top of regular search results, reaching many more people by default.
The bigger worry is its own user base. Power users still appreciate the citations and the Comet browser. However, paying users are upset. The company was caught quietly replacing more expensive models with cheaper ones and then sharply limited usage with little warning. A privacy lawsuit and copyright lawsuits from major publishers add to this situation.
For a product whose main asset is trust, this combination is risky.
MY TAKE
Will Perplexity still matter in five years?
Not if it keeps going this way, and I say that as someone who has been using Comet since the beta. The browser is really good. However, Claude already performs the same tasks through a simple Chrome extension. Google now provides AI answers directly in search. Most people still open Chrome out of habit. Comet is a better product fighting against a default like Chrome, and defaults usually win. The one thing people still trust Perplexity for is its citations and clean answers they can verify. That is its only real advantage, and lately it has been spending trust to save money through quiet model swaps and sudden limit cuts while lawsuits pile up. Anyone can now create cited answers, and the large companies already have. The product is good, but the position is weak.
Here is the deeper problem: nothing keeps anyone in. Perplexity has no switching cost. A consumer can switch to ChatGPT, Gemini, or plain Google in seconds and lose nothing, no saved work, no locked-in habit, no reason to stay. The simplest solution is the one OpenAI and Anthropic are already pursuing: enterprise. Companies pay much more than $20 a month. Once Perplexity is integrated into their data, tools, and daily workflows, leaving becomes costly. That is true lock-in, based on the trust it already has. If it focuses there and builds something new on top of it, it will endure. If it keeps competing with Google on Google's terms, it will become just a footnote.
04. ONE THING TO WATCH
Watch for Perplexity's next major launch: does it provide people with a real reason to stay maybe with memory, saved work, true lock-in or is it just another feature that Google can copy by the end of the year?
That's the teardown.
Reply and tell me your feedback and which AI startup you want torn apart next. I read every one.
- Abhishek
P.S. If you found this useful, forward it to a founder who's tired of the hype.