Perplexity Revenue: How It Grew to $750M and 100M+ Users

Perplexity started with about 3,000 queries on its first day in 2022.

By May 2025, it was handling 780 million queries a month. By 2026, Perplexity said it had more than 100 million monthly users across its products, while Reuters reported that its annualized revenue had passed $750 million.

The growth came in stages.

First, Perplexity built a search product people wanted to tell others about. It gave direct answers, searched the live web, and showed citations next to its claims.

Then it removed as much friction as possible. No login was required for the first search. Follow-up questions encouraged another search. Shared answers brought new people into the product.

Once that organic engine worked, Perplexity added distribution: mobile apps, telecom deals, device partnerships, free Pro bundles, and eventually its own browser.

Its growth strategy can be summarized like this:

better search experience → word of mouth → repeated searches → partnerships → default distribution → enterprise and agent revenue

That sequence matters.

Perplexity did not start by buying hundreds of millions of users.

It built something people kept using first.

Perplexity's growth at a glance

The numbers do not all measure the same thing. Queries, downloads, monthly users, and annualized revenue should not be treated as interchangeable.

But together they show how quickly Perplexity moved from a small AI search experiment into a large consumer and enterprise product.

1. Perplexity launched at exactly the right time

Perplexity was founded in August 2022.

Its answer engine launched in December 2022, only days after ChatGPT had shown millions of people that asking questions to an AI could be useful.

But early ChatGPT had an important weakness: it did not search the live web and did not reliably show where its answers came from.

Perplexity took a different approach.

Its first public version combined GPT-3.5 with Bing search results. A user could ask a question in natural language and get a short answer based on current web pages.

Most importantly, the answer included citations.

The company's December 2022 Hacker News launch described the idea simply: use a traditional search engine to find information, then use an LLM to summarize the results and cite the sources.

No login was required.

No ads appeared around the answer.

Users did not have to open ten links and piece the answer together themselves.

That gave Perplexity a clear position between Google and ChatGPT.

Google gave you links.

ChatGPT gave you answers.

Perplexity tried to give you answers with evidence.

That was the wedge.

2. Citations made AI search easier to trust

Perplexity's biggest early product decision was not inventing a new model.

It was showing sources.

This mattered because AI answers were already developing a trust problem. Large language models could produce confident answers even when they were wrong.

Perplexity gave users something they could check.

Its Series A announcement in 2023 said the company wanted to provide citations with every answer so users could verify where the information came from. By that point, Perplexity said it had reached 2 million monthly active users in four months.

Citations did not guarantee that every answer was correct.

But they changed how the product felt.

Instead of asking:

Can I trust this AI?

A user could ask:

Where did it get this from?

That small change helped make Perplexity useful for research, coding, school, work, news, travel, and other searches where the source mattered.

It also gave people a simple reason to recommend the product:

"It's like ChatGPT, but it searches the web and cites its sources."

Good products are easier to spread when their advantage fits into one sentence.

3. Perplexity removed friction from the first search

The original Perplexity did not force users to create an account before trying it.

You landed on the page.

You typed a question.

You got an answer.

That sounds obvious now, but it was an important growth choice.

For a new product, every extra step before the "aha" moment gives people another chance to leave.

Perplexity made the useful action the first action.

The team also focused on queries, not registrations, as a key measure of product usage.

The goal was not simply to get somebody onto the website.

The goal was to get them searching.

Perplexity then designed the product to turn one search into another.

Autosuggest made questions easier to write.

Related questions appeared under answers.

Threads let users continue researching a topic without starting over.

The result was a simple loop:

ask → get an answer → see a related question → ask again

This mattered for retention.

A product like Perplexity becomes much harder to forget once searching it becomes a habit.

4. The activation goal became multiple searches, not one signup

Perplexity's growth team later became more precise about this behavior.

Growth leader Raman Malik has discussed getting users to make several searches during their first experience with the product. Three searches became an important activation point.

This is a much better goal for a search product than "create an account."

An account does not prove somebody understands the product.

Several searches do.

It also explains why some of Perplexity's later marketing campaigns were designed around usage instead of simple awareness.

The company was not only trying to make people know Perplexity existed.

It wanted them to build the habit.

5. Word of mouth became the main acquisition channel

Perplexity's early growth was unusually organic.

People tried the product, got a useful answer, and told somebody else.

In a 2025 interview, Malik said word of mouth accounted for roughly 80% of acquisition in some markets.

That is an important detail in the Perplexity story.

The company later spent heavily on partnerships, promotions, creators, sponsorships, and other forms of distribution.

But those channels came after people were already recommending the product.

Perplexity had product-led growth before it had large-scale paid distribution.

The difference matters.

Partnerships can put a product in front of millions of people.

They cannot make the product worth returning to.

6. Sharing turned answers into distribution

Perplexity answers could be shared with a link.

That meant users were not limited to sending screenshots.

They could send someone the full answer, along with its sources and research trail.

Those links moved through X, LinkedIn, Slack, WhatsApp, email, iMessage, Reddit, and group chats.

Each shared answer became a small landing page for Perplexity.

The recipient did not need to understand what Perplexity was first.

They could read the answer and then try another question.

Perplexity later expanded this behavior with Threads and Pages.

Pages, launched in 2024, let users turn research into a cleaner public page that could also appear on the open web.

This was not a perfect viral loop. Most people who search do not automatically invite somebody else.

But it gave Perplexity an organic distribution surface that normal chatbot conversations did not have.

7. Aravind Srinivas became one of Perplexity's best distribution channels

Perplexity also benefited from founder-led marketing.

CEO Aravind Srinivas posted constantly.

He announced features.

He replied to users.

He debated competitors.

He talked about Google.

He shared usage numbers.

He responded to criticism.

He explained what the team was building next.

That kept Perplexity inside the wider conversation around AI search.

The strategy worked especially well because the company shipped often.

Every product release gave Srinivas something new to talk about.

And every post about a release brought people back into the product.

The loop looked like this:

ship → announce → users try it → users discuss it → team learns → ship again

Perplexity turned shipping into marketing.

8. New products kept giving users a reason to come back

Perplexity did not leave the original search experience untouched.

It kept adding new reasons to use it.

Over time, the company launched products and features for deeper research, file uploads, multiple AI models, finance, shopping, enterprise search, Pages, Spaces, APIs, Comet, and Computer.

Not every launch needed to become a huge business by itself.

Each one did something else that mattered:

it put Perplexity back into the news.

For a fast-moving AI company, product velocity became part of the distribution strategy.

Users kept hearing about Perplexity because Perplexity kept changing.

9. Perplexity expanded from a website into a mobile product

The next step was moving beyond the desktop website.

Perplexity launched its iPhone app in April 2023.

At the time, the company said almost half of its users were already accessing Perplexity from mobile web.

The app made repeated use easier.

Voice search made the product useful in more situations.

Notifications and a permanent app icon increased the chance that users would remember Perplexity instead of automatically opening Google.

But building an app still left the same hard problem:

people already had default search habits.

Perplexity needed distribution.

10. Partnerships became Perplexity's second growth engine

Once the product had proven that people wanted it, Perplexity started borrowing distribution from companies that already had large audiences.

It partnered with telecom operators, device makers, subscription services, and other consumer platforms.

The logic was simple.

Google owned huge distribution through Chrome, Android, and default search placement.

Perplexity did not.

So instead of building every distribution channel from scratch, it partnered with companies that already controlled one.

Deals included telecom groups such as SK Telecom, SoftBank, Deutsche Telekom, and Airtel, along with product and device partnerships.

Some promotions gave partner customers months or even a year of Perplexity Pro for free.

This dramatically lowered the cost of trying the paid product.

Perplexity could give away access because someone receiving a free account did not create the same cost as giving away a physical product.

If the user never used it, the cost was limited.

If they did use it, Perplexity had a chance to create a new habit.

11. Airtel became Perplexity's biggest growth experiment

The clearest example came in India.

In July 2025, Airtel offered one year of Perplexity Pro for free to its 360 million customers.

The effect was immediate.

According to Sensor Tower data reported by TechCrunch, Perplexity recorded 5.9 million downloads in India during July 2025.

That was a 625% increase from the previous month.

Over the seven months when people could redeem the Airtel offer, Perplexity recorded an estimated 56 million downloads in India.

Monthly active users eventually peaked around 22 million.

The promotion ended.

Downloads then fell sharply.

But usage did not fall all the way back to where it started.

By July 2026, Sensor Tower estimated that Perplexity still had nearly 14 million monthly active users in India, compared with about 2.6 million a month before the promotion.

That makes Airtel one of the best examples of Perplexity's later growth strategy.

The company used a partner to get huge distribution.

Then it relied on the product to retain part of that audience.

The partnership did not make every free user permanent.

It did leave Perplexity with a much larger user base than it had before.

12. Perplexity used stunts to make people actually use the product

Perplexity also became good at attention-grabbing marketing.

One of its best campaigns happened during the 2025 Super Bowl.

Instead of buying a traditional Super Bowl commercial, Perplexity offered $1 million to a user who downloaded the app and asked at least five questions during the game.

The important part was the second requirement.

Perplexity did not reward a download alone.

To enter, users had to use the product repeatedly.

According to Appfigures data reported by TechCrunch, daily installs increased from around 30,000 to 45,000, a rise of roughly 50%.

The campaign matched Perplexity's activation strategy.

A normal ad tells somebody a product exists.

This campaign made them experience it.

13. Perplexity moved closer to being the default

Partnerships solved part of the distribution problem.

The next step was more ambitious.

Perplexity wanted to control the place where searches begin.

That led to Comet.

Perplexity launched its AI browser in limited release in July 2025 and made it free worldwide in October.

Why build an entire browser?

Because a browser creates many more chances to use Perplexity.

Instead of visiting Perplexity's website, a user could interact with it from the search bar, a new tab, a sidebar, or the page they were already reading.

Srinivas described the opportunity as "infinite retention."

In May 2025, Perplexity was already processing 780 million queries per month, growing more than 20% month over month according to Srinivas. He argued that putting Perplexity inside the browser could increase both queries per user and the number of users.

Comet was therefore not just another product launch.

It was a distribution strategy.

Google had Chrome.

Perplexity wanted its own place where user intent started.

14. Device integrations pushed the same strategy even further

Perplexity also moved deeper into hardware distribution.

Samsung integrated Perplexity into its Galaxy S26 phones in 2026.

Perplexity said its app would come preloaded on every Galaxy S26, with search and reasoning integrated with Bixby and Samsung apps.

This is a very different growth model from the company's first year.

Early Perplexity depended on someone hearing about the product and visiting its website.

The later strategy tries to put Perplexity in places where the user is already searching.

That shift can be described simply:

2022–2023: make people choose Perplexity

2024–2026: put Perplexity where people already are

15. Enterprise turned consumer adoption into another growth loop

Perplexity also expanded from individual search into work.

It launched Enterprise Pro in 2024.

Companies could use Perplexity with features such as user management, security controls, internal knowledge, and enterprise data.

This opened another growth path.

An individual could discover Perplexity at home.

They could start using it for work.

Their coworkers could see it.

Eventually, the company could buy Perplexity for the wider organization.

Today Perplexity says more than 50,000 organizations use its enterprise product.

Consumer adoption helped introduce the product.

Enterprise turned some of that usage into larger contracts.

16. Perplexity Computer changed the revenue model

The company then moved beyond answering questions.

In February 2026, it launched Perplexity Computer, an agent designed to perform longer tasks using multiple models and tools.

That changed the economics.

Search is usually a short interaction.

An agent can perform much larger pieces of work.

And larger tasks can support usage-based pricing.

The Financial Times reported that Perplexity's monthly revenue jumped roughly 50% in March 2026, taking annualized revenue above $450 million. The report linked much of the acceleration to Computer and the shift toward agentic products.

By August 2026, Reuters reported that annualized revenue had increased from below $250 million at the beginning of the year to more than $750 million. Perplexity declined to confirm the number.

The figure is an annualized run rate, not audited annual revenue.

But the direction is important.

Perplexity first used search to build a large audience.

Then it added higher-value products that could earn more from power users and companies.

Perplexity's growth strategy in one chart

The company effectively moved through four stages:

The order is the important part.

Perplexity did not begin with the Airtel deal.

It began with a product that already had organic pull.

That made large distribution deals much more valuable later.

What did not work as well

Perplexity's growth was not a straight line.

The company tested paid acquisition, creator campaigns, newsletters, ads, and other channels.

Some performed poorly.

Its growth team has said broad influencer campaigns required too much coordination and did not work as well as deeper relationships with a smaller number of creators.

Perplexity also experimented with sponsored questions.

It eventually moved away from advertising after concluding that ads could weaken trust in the answers. The Financial Times reported in 2026 that Perplexity had phased out advertising and was focusing more heavily on subscriptions and paid agent products.

That decision fits the core product.

If users believe money can influence the answer, citations become less valuable.

Trust is unusually important when the product itself is giving people information.

The biggest risk in Perplexity's growth model

The same thing that helped Perplexity grow has also created its biggest problem.

Perplexity answers questions using information from the web.

That saves users from opening many of the pages the information came from.

Publishers have argued that this reduces their traffic while Perplexity captures the value of their reporting.

The company has faced complaints and lawsuits from media companies and other publishers over crawling, attribution, and copyright.

There is also a product problem.

A citation makes an answer look trustworthy, but a citation does not guarantee the source actually supports every claim next to it.

So Perplexity has to solve a difficult problem as it grows:

How do you make answers faster without making the open web that supplies those answers less sustainable?

That question remains unresolved.

Why Perplexity grew so quickly

Perplexity's growth did not come from one marketing trick.

Several things compounded.

It entered the market when millions of people were suddenly interested in conversational AI.

It found a simple product difference: answers with sources.

It made the first search almost frictionless.

It optimized for repeated searches instead of signups.

Its users recommended it.

Its founder kept the company visible.

The team shipped constantly.

Then, only after the product had organic pull, Perplexity started using larger distribution channels.

Telecom partnerships put it in front of millions of people.

Airtel put it in front of hundreds of millions.

Comet gave it control over the browser.

Device deals put it closer to being a default.

Enterprise and Computer increased how much valuable users could spend.

That is the full Perplexity growth loop:

build something worth searching with → turn searching into a habit → let users spread it → borrow distribution → own more distribution → monetize deeper usage

What startups can learn from Perplexity

Make the product easy to explain

Perplexity's early pitch could fit into one sentence:

Ask a question, get an answer, and see the sources.

If a user cannot quickly explain why your product is different, word of mouth becomes harder.

Optimize for the action that creates retention

Perplexity did not stop at signups.

It wanted people to search several times.

Every startup should know the action that separates users who leave from users who stay.

Remove friction before the aha moment

The original Perplexity did not require a login before users could experience the main product.

Show people the value first.

Ask for commitment later.

Use marketing to drive product usage

The Super Bowl campaign did not simply ask people to download Perplexity.

It asked them to make five searches.

The best campaigns teach people the product while acquiring them.

Add distribution after the product works

Airtel's 360 million customers look like the biggest part of the story.

They are not.

The important part happened earlier.

Perplexity had already shown that people would discover it, use it repeatedly, and recommend it without a giant telecom partner.

Distribution amplified something that already worked.

Own more of the user journey over time

Perplexity started as a website.

Then it built apps.

Then partnerships.

Then device integrations.

Then a browser.

Each move brought it closer to the moment when a user decides to search.

That reduces dependence on somebody else's platform.

Getting cited in AI answers

Perplexity is now one of the places buyers ask for recommendations, and every answer shows its sources. So are ChatGPT and Google's AI Overviews. Okara's GEO Agent tracks up to 15 of your buyers' prompts in ChatGPT and Google AI Overviews and shows which competitors get cited instead of you. For the other half of Perplexity's playbook, a founder who shows up every day, the X Agent drafts posts in your voice.

Frequently asked questions

How did Perplexity grow?

Perplexity grew first through product-led growth and word of mouth. Its cited web answers, no-login onboarding, follow-up questions, sharing, and founder-led marketing helped it acquire early users. It later accelerated growth through mobile apps, telecom bundles such as Airtel, device partnerships, its Comet browser, and enterprise products.

What was Perplexity's main growth strategy?

Perplexity's main early strategy was to build a search product people recommended organically. Growth leader Raman Malik said word of mouth accounted for around 80% of acquisition in some markets in 2025. Large partnerships became more important after the product had already reached millions of users.

How many queries does Perplexity get?

CEO Aravind Srinivas said Perplexity handled 780 million queries in May 2025, up from roughly 3,000 queries on its first day in 2022. The company was growing query volume by more than 20% month over month at the time. Perplexity has since expanded into browsers, agents, APIs, and enterprise products, making newer usage figures harder to compare directly.

How many people use Perplexity?

The Financial Times reported in March 2026 that Perplexity had more than 100 million monthly users across its products. Other third-party estimates of its standalone search and app usage are lower because they measure a narrower set of products.

How much revenue does Perplexity make?

Reuters reported in August 2026 that Perplexity's annualized revenue had passed $750 million, up from less than $250 million at the start of the year. Perplexity declined to confirm the figure. Annualized revenue is a run rate based on current revenue, not the same as audited revenue for a completed year.

Why did Perplexity become popular?

Perplexity became popular because it combined the conversational experience of an AI chatbot with live web search and visible citations. Users could get a direct answer without opening many search results while still seeing the sources behind the response.

What made Perplexity different from ChatGPT?

At launch, ChatGPT mainly answered from information contained in its model. Perplexity searched the live web and attached citations to its answers. That made it especially useful for questions where freshness and source verification mattered.

What was the Airtel and Perplexity partnership?

In July 2025, Airtel offered a free year of Perplexity Pro to 360 million customers in India. Perplexity recorded 5.9 million Indian downloads that month, up 625% from the previous month. Sensor Tower estimated that monthly active users remained far above their pre-promotion level a year later.

What is Perplexity's valuation?

Perplexity's last reported valuation was about $20 billion, set in September 2025. In August 2026, Reuters, citing The Information, reported that Nvidia was in talks to invest in a new round that would value Perplexity at more than $30 billion. That round had not been confirmed at the time of writing.

The bottom line

Perplexity's growth story is easy to misunderstand.

The telecom deals were huge.

The $1 million promotions got attention.

Comet gave the company its own browser.

But none of those created the original demand.

The most important growth decision happened in 2022.

Perplexity made AI search feel more useful by giving people a direct answer and showing where that answer came from.

Then it made the experience easy to try and easy to repeat.

Everything else amplified that behavior.

Perplexity went from roughly 3,000 queries on its first day to 780 million queries in a month by May 2025. Three years later, it was reporting more than 100 million users across its products and was generating a reported annualized revenue run rate above $750 million.

The lesson is simple:

product creates the pull. distribution multiplies it.

Sources

• Perplexity — Series A funding and mobile app announcement, April 2023
• Hacker News — Perplexity's original December 2022 launch
• Reuters — Nvidia discusses Perplexity investment at $30 billion-plus valuation, August 2026
• TechCrunch — "Perplexity received 780 million queries last month, CEO says," June 2025
• TechCrunch — "Perplexity's Super Bowl 'tweet' increased app installs by 50%," February 2025
• TechCrunch — "Perplexity sees India as a shortcut in its race against OpenAI," July 2025
• TechCrunch — "Perplexity's free AI offer left it with millions more users in India," August 2026
• Perplexity — Comet launch and worldwide release announcements
• Perplexity — Samsung Galaxy S26 integration announcement
• Financial Times — Perplexity revenue and AI agent reporting, March 2026
• Reuters — Perplexity reported annualized revenue and valuation discussions, August 2026