Perplexity’s share of the AI chatbot market has fallen from roughly 6.6% in 2025 to somewhere between 1.3% and 2% in 2026, according to Similarweb data cited by multiple industry trackers. At the same time, its annualized revenue climbed from about $63 million at the end of 2024 to roughly $450–500 million by April 2026 — a nearly eightfold increase. The two trends aren’t contradictory. Perplexity has deliberately shifted from chasing casual chatbot users toward selling expensive, usage-based AI “agent” tools to power users and enterprises, and that smaller, higher-paying base is now worth more than a much larger free audience ever was.
Three years ago, Perplexity was being described as a genuine threat to Google — a startup confident enough to offer $34.5 billion for Google’s Chrome browser despite being worth less than that itself. Today, its slice of the AI chatbot traffic pie is smaller than it’s ever been. Yet its revenue chart looks nothing like a company in decline. Understanding how both things are true at once says a lot about where the AI search business is actually headed in 2026.
How Big Was Perplexity’s Market Share Collapse?
Perplexity‘s chatbot market share peaked in 2025 and has fallen sharply since. Multiple analytics sources tracking AI chatbot web traffic put Perplexity’s global share at around 2.0% in early 2026, sliding further to roughly 1.3% by May, according to Similarweb data referenced by industry trackers — making it the smallest of the seven major AI assistants tracked, behind ChatGPT, Gemini, DeepSeek, Grok, and Meta AI. On US mobile specifically, its daily-active-user share reportedly fell from around 6% to 2% in just five months.
Website visits tell a similar story: Perplexity’s traffic peaked at roughly 240 million visits in November 2025 and has settled into the 170–180 million range since — a meaningful pullback, even as the broader AI assistant category keeps growing.
So Why Is Perplexity’s Revenue Still Climbing?
This is the part that looks counterintuitive until you see the pricing shift behind it. Perplexity’s annualized recurring revenue (ARR) moved from approximately $63 million at the end of 2024, to $200 million by September 2025, to more than $450 million in March 2026, with independent estimates from research firm Sacra putting it near $500 million by April 2026. Management’s own internal target, according to reporting from the Financial Times, is $656 million in ARR by the end of 2026.
Three shifts explain the disconnect between shrinking traffic and growing revenue:
- Perplexity dropped advertising entirely in February 2026, betting that a fully subscription-based, ad-free answer engine would build more trust with the paying users it had left — and revenue reportedly jumped roughly 50% the following month.
- It moved from flat-rate to usage-based pricing. A casual user asking a handful of questions a day was never going to generate much revenue at $20/month. A power user running dozens of complex, multi-step “agent” tasks a day burns through far more compute — and pays for it.
- It shifted its center of gravity from consumers to enterprises, where a single company license can be worth more than 200 individual subscriptions combined.
From a $9 Billion Longshot to a $22.6 Billion Bet
Perplexity’s funding history mirrors this arc. A $3.1 million seed round in 2022 grew into a $9 billion valuation by December 2024, roughly $18–20 billion through mid-to-late 2025, and $22.6 billion following a Series E-6 round in January 2026. Backers have included SoftBank, Nvidia, Jeff Bezos, Accel, IVP, and Databricks. Against roughly $450–500 million in ARR, that valuation implies a revenue multiple in the 45–50x range — rich, but down sharply from the 100x+ multiple the company commanded at its September 2025 raise, suggesting investors are pricing in more caution about how long the current growth rate can hold.
The company’s boldest and strangest swing during this period was a genuinely audacious one: in August 2025, Perplexity submitted an unsolicited $34.5 billion all-cash bid to buy Google’s Chrome browser, timed to a US antitrust ruling that raised the possibility Google could be forced to divest it. The offer was nearly double Perplexity’s own valuation at the time, and analysts broadly dismissed it as unlikely to succeed — Chrome commands roughly two-thirds of the global browser market and sits at the center of Google’s advertising business. Google never agreed to sell, but the bid made clear how seriously Perplexity was thinking about distribution, not just search quality, as its actual bottleneck.
The Real Problem: Google’s Distribution Moat
Perplexity’s founding insight in 2022 was genuinely sharp: ChatGPT could explain things but had no live access to the internet, while Google could search the entire internet but couldn’t hold a conversation about what it found. Perplexity combined a search backend with a language model and generated direct, cited answers — a genuinely new experience that took it to 2 million users within four months of launch.
The problem is that this gap didn’t stay open for long. OpenAI launched ChatGPT Search in October 2024. Google rolled out AI Overviews and expanded them broadly by May 2025. The exact feature that made Perplexity special became a standard checkbox in nearly every major chatbot. And distribution — not answer quality — decided what happened next.
Google’s advantages here are structural, not just technical. Chrome is the world’s largest browser. Android is the world’s largest mobile operating system. Google is the default search engine in Safari, reportedly under a deal worth around $20 billion a year to Apple. Even Microsoft, despite spending billions building Bing, owning Windows and Edge, and investing heavily in OpenAI, has never meaningfully dented Google’s search share. Against that kind of distribution, a three-year-old startup was never going to out-market its way to Google-level reach through search results alone — no matter how good the underlying product was.
The Pivot: Perplexity Computer and the Agent Bet
Rather than continuing to fight Google head-on for search traffic, Perplexity’s 2026 strategy has been to build something Google, OpenAI, and Anthropic hadn’t fully shipped yet: a general-purpose AI agent that actually executes tasks rather than just answering questions.
On February 25, 2026, Perplexity launched Perplexity Computer, a cloud-based system that coordinates 19 different AI models to plan and execute multi-step workflows — research, spreadsheets, presentations, even functioning web apps — inside an isolated cloud environment, priced at $200/month under the company’s Max tier. Two weeks later, at its inaugural “Ask 2026” developer conference, Perplexity extended the product into Computer for Enterprise, adding Slack and Microsoft Teams integration, more than 400 app connectors (Snowflake, Salesforce, HubSpot, GitHub, and others), SOC 2 Type II compliance, and enterprise pricing around $325 per seat per month. The company said more than 100 enterprise customers signed on almost immediately.
This single decision reframed Perplexity’s entire business model:
| Old Model (Search) | New Model (Computer / Agents) | |
|---|---|---|
| Pricing | Flat $20/month, generous free tier | Usage-based; heavy tasks consume more credits |
| Buyer | Individual consumer | Individual power user + enterprise seat licenses |
| Revenue per customer | Low and capped | Uncapped — scales with task complexity |
| Competitive comparison | Compared directly to Google, ChatGPT | Compared to Microsoft Copilot, Salesforce, enterprise software stacks |
| What wins | Search relevance and speed | Trust, integrations, and workflow automation |
A single enterprise client buying 100 seats at roughly $40 per user per month can be worth close to $48,000 a year to Perplexity — a figure the company would need around 200 individual annual Pro subscribers to match. Free users, no matter how many of them there are, generate essentially nothing directly. That arithmetic is the entire explanation for why a shrinking market-share number and a growing revenue number can appear on the same earnings snapshot.
Is Perplexity Actually Winning?
Not against the traffic numbers — not yet, and maybe not ever, if traffic share is the yardstick. But traffic was never really the business. A company generating close to $500 million in annualized revenue from roughly 1–2% of AI chatbot web traffic is, by ordinary software-industry standards, extracting far more value per user than its market-share ranking would suggest.
The bigger risk is competitive, not financial. Perplexity’s head start in agentic AI tools is real but narrow. Elon Musk’s xAI has since launched a comparable product (Grok’s agent tooling), Anthropic has shipped Claude Cowork, and OpenAI has rolled out ChatGPT Work — both running tasks directly on a user’s own machine rather than in an isolated cloud sandbox, which is a meaningfully different (and for some enterprise buyers, more trusted) architecture. Perplexity is the first mover in this specific race, but first movers in AI have not historically stayed ahead for long once the giants decide the category matters.
Frequently Asked Questions
Is Perplexity AI losing market share in 2026?
Yes. Perplexity’s share of global AI chatbot web traffic fell from roughly 6.6% in 2025 to somewhere in the 1.3%–2% range by mid-2026, according to Similarweb-based tracking, making it the smallest of the seven major AI assistants by traffic share.
Why is Perplexity’s revenue growing if its market share is shrinking?
Perplexity shifted from a flat-rate, ad-supported consumer product to a usage-based, subscription-only model focused on power users and enterprise customers. A much smaller but higher-paying base — especially enterprise seats on its Computer agent product — now generates far more revenue than a larger free user base ever did.
What is Perplexity Computer?
Perplexity Computer, launched February 25, 2026, is a cloud-based AI agent that coordinates 19 different AI models to plan and complete multi-step tasks — research, spreadsheets, presentations, and working web apps — with minimal human input. An enterprise version, launched weeks later, adds Slack, Microsoft Teams, and 400+ business app integrations.
How much is Perplexity worth in 2026?
Perplexity was valued at approximately $22.6 billion following a Series E-6 funding round in January 2026, up from about $9 billion in December 2024, on annualized revenue of roughly $450–500 million.
Did Perplexity really try to buy Google Chrome?
Yes. In August 2025, Perplexity made an unsolicited $34.5 billion all-cash offer to acquire Google’s Chrome browser, timed to a US antitrust ruling that raised the possibility Google could be forced to sell it. Google did not agree to the offer.
Is Perplexity’s growth strategy sustainable?
It depends on whether the enterprise agent market keeps expanding as projected — one industry estimate puts the global agentic AI market growing from roughly $9 billion in 2026 to $139 billion by 2034 — and whether Perplexity can defend its head start against similar products now shipped by OpenAI, Anthropic, and xAI.
Read More:
