Perplexity AI Features and Statistics 2026: What You Can Actually Do

Perplexity AI processed 780 million search queries in May 2025. That was up from about 230 million in mid-2024, so it roughly tripled in under a year. By mid-2026 the platform is estimated to handle 1.2 to 1.5 billion queries a month.

But raw numbers do not tell you what the platform actually does. In 2026 Perplexity has Pro Search for deep research, Spaces for organizing work, a Model Council that compares answers from several models side by side, and the free Comet browser that reached number three on the US App Store within 48 hours of its iOS launch.

This guide covers what Perplexity can do, how people use it, and the growth stats that show why developers are switching from Google. Whether you are a developer, researcher, business owner, or product builder, this data shows why Perplexity matters and where it is headed next. Full breakdown below.

Transform your tech career. Join Index.dev and get matched with top companies that value innovation. Index.dev connects the top 1% of senior engineers from LATAM and CEE with global teams, with matches in 48 hours.

5 Key Takeaways

• Perplexity is estimated to run 1.2 to 1.5 billion queries a month by mid-2026, up from about 230 million a month in August 2024.
• The platform reports over 100 million monthly active users across all products in 2026, per Sacra’s April 2026 analysis, and now carries a ~$20 billion valuation after a $200 million round in September 2025.
• Perplexity’s edge is cited answers from live web sources, not a page of links. Pro Search reads 300+ sources per query and the new Model Council compares GPT-5.2, Claude 4.6, and Gemini 3.1 Pro side by side.
• The free Comet browser went worldwide in October 2025 and hit number three on the US App Store within 48 hours of its iOS launch in March 2026.
• Citation-backed search is a research tool, not a code generator. For verifying technical claims it beats Google link lists, which is why teams pair it with strong AI engineers who can build on top of it.

What Perplexity AI Actually Does in 2026

Here is what you get with Perplexity right now.

Core Features (2026)

• Pro Search and Deep Research: searches hundreds of sources with clarifying follow-ups for current and deeper context.
• Comet AI browser: a free, AI-augmented Chromium browser that summarizes pages, manages tasks, and automates multi-step workflows.
• Background Assistant (Max plan): a task assistant that runs jobs in the background and helps across apps.
• Voice search and mobile assistant: voice queries and assistant features live in the Perplexity mobile app.
• Memory and personalization: remembers past threads and interests to improve context and answers across models.
• Multi-model support: access several AI models for different needs, such as reasoning, coding, and search.
• File upload and analysis: upload docs (PDF, Word, and more) for direct AI querying and extraction.
• Shopping and commerce tools: integrated shopping and purchase flows powered by PayPal inside conversational search.

While Comet simplifies how we consume information, the rise of AI browser automation through tools like Skyvern lets users turn those insights into actions.

Where You Can Use It

• Web app at perplexity.ai
• iOS and Android mobile apps
• Desktop apps for Windows and macOS via the Comet and Perplexity clients

Language and Regional Support

• Works in 40+ languages. Regional bundles and pricing are available in select markets, such as carrier deals, though some telecom perks may change over time.

Pricing (2026)

• Free tier: basic AI search and Comet
• Pro plan: advanced models, document analysis, and priority tools (about $20 per month)
• Max plan: full enterprise features including the Background Assistant and Comet Plus (about $200 per month)

Now let us get into the numbers that explain why this matters.

1. Search Growth and Platform Scale

Perplexity AI is scaling its ability to answer queries at a fast pace.

• By mid-2026, Perplexity is estimated to process 1.2 to 1.5 billion search queries per month, driven by Pro Search, Comet usage, and mobile growth.
• Weekly query volume is now estimated at 180 to 220 million searches, more than double early-2025 levels.
• Daily usage has crossed 35 to 45 million queries per day, up from about 30 million per day in 2025.
• For comparison, Perplexity handled about 230 million monthly queries in August 2024, so usage has grown roughly 5 to 6 times in under two years.
• The company has not publicly confirmed hitting 1 billion weekly queries, but internal growth trends point toward that scale over time.
• Total queries processed across 2025 and 2026 combined are estimated to exceed 6 to 7 billion, far past all pre-2024 usage.
• For context, in 2022 Perplexity handled only a few thousand queries per day. By 2026 it runs at tens of millions of daily queries, one of the fastest scale curves in consumer AI.

This fast growth proves the need for strong backend systems, scalable infrastructure, and AI engineers who can support real-time, citation-heavy performance.

2. Global Visitors and Geographic Reach

Perplexity AI has rapidly expanded its global presence and attracted millions of users worldwide. Here is how its growth looks.

• In May 2025, Perplexity AI recorded 153 million website visits, making it one of the most-visited AI platforms globally.
• In March 2024, the platform had 52.4 million visits, so traffic grew by 191.9% in just over a year.
• Perplexity AI now holds a global website rank of 268, showing strong visibility.
• The platform is available in 238 countries, proving wide international reach.
• It supports 46 languages, letting users access the tool in their native language.
• On average, around 2 million people visit the platform each day.
• In June 2025, the platform received 141.32 million visits, holding strong traffic even after May’s peak.
• Across 2026, Perplexity reports over 100 million monthly active users across all products combined, per Sacra’s April 2026 analysis citing CEO Aravind Srinivas.

This kind of global reach highlights the need for developers, data engineers, and product teams who can support scalable, multilingual, and globally distributed systems.

3. Where Are the Users Coming From?

Perplexity’s user base is diverse, but several countries stand out.

• The United States is the top traffic source, with over 23 million visits, mostly from desktop users.
• India contributes nearly 12 million visits, with slightly higher mobile usage than other countries.
• Germany, South Korea, and Japan all show a strong preference for desktop access, with over 87% of users on computers.
• South Korea has the highest desktop usage at 92.32%, followed closely by Japan and Germany.

Indonesian Market: The Telkomsel Partnership

Perplexity partnered with Telkomsel in May 2025. Telkomsel is Indonesia’s biggest mobile operator with 178 million subscribers.

The deal: Indonesian users can buy Perplexity Pro bundled with data plans. Prepaid customers pay 35,000 rupiah (about $2.30) for 3 months of Pro plus internet quota.

Why it matters: This was Perplexity’s first telecom partnership in Southeast Asia. The app fully supports Bahasa Indonesia for both interface and search queries.

So Indonesian developers can search in their native language and get results from both local and international sources. The bundled pricing makes Pro accessible to far more users.

4. Device Usage: Desktop vs Mobile

Perplexity AI users behave differently depending on their location and preferences.

• In March 2025, 133.3 million visits came from desktops, while 26.4 million came from mobile devices.
• In June 2025, 82.51% of visits were from desktop and 17.49% from mobile.
• In the United States, 59.69% of users access Perplexity on desktop.
• In Indonesia, 92.53% of users access it via mobile, showing different preferences by region.

These patterns mean developers need to build for both desktop and mobile, especially in mobile-first regions.

5. How Long Do Users Stay?

Perplexity’s engagement metrics show that users visit often and also stay longer than on typical sites.

• On average, each visitor spends 23 minutes and 10 seconds per session, far longer than the average for search platforms.
• Users view around 4.64 pages per visit, exploring multiple queries in one session.
• The bounce rate is 42.19%, so most visitors stay and engage instead of leaving after one page.
• 85% of users return after using it once, a sign of strong satisfaction.

• 90% of users revisit within 30 days, showing high retention and habit.

These stats confirm that Perplexity delivers an engaging experience that keeps users coming back.

6. Who Are the Users?

Perplexity AI attracts a young, tech-focused audience that values speed, accuracy, and productivity.

• The largest share of users sits in the 18 to 34 age group, at 53.24% of the base. This includes students, early-career developers, and professionals.
• 20.22% of users are 18 to 24, while 33.02% are 25 to 34, showing strong adoption among Gen Z and Millennials.
• The next brackets: 19.67% aged 35 to 44, 13.63% aged 45 to 54, 8.26% aged 55 to 64, and 5.20% aged 65 and above.
• By gender, 60.19% of users are male and 39.81% female, a slightly higher share of men but strong engagement from both.

This young, tech-savvy audience drives demand for AI products that are fast, intuitive, mobile-friendly, and developer-oriented.

7. Where Does the Traffic Come From?

Perplexity AI benefits from a healthy mix of direct, organic, and social traffic, a sign of strong brand presence.

• Direct visits account for 77.8% to 82% of traffic, a sign of high brand recall and repeat usage.
• Organic search contributes 11.65% to 16.6% of traffic, so Perplexity keeps attracting new users through search engines.
• Social media generates around 7.5% to 10% of total traffic, expanding reach.

Social Media Breakdown

• YouTube leads with 50.95% of social traffic, driven by explainer videos and feature demos.
• Facebook contributes 8.26%, Twitter (X) 7.67%, and LinkedIn 7.53%.
• WhatsApp accounts for 6.98%, a sign of peer-to-peer sharing.
• Other platforms make up the remaining 18.62% of social traffic.

Together, this multi-channel presence reflects strong engagement across both direct usage and external discovery.

8. Platform Performance and Accuracy

Perplexity AI has built a reputation for fast, well-cited answers, which makes it a go-to for reliable information.

• Perplexity reports a 95% accuracy rate on its search results, based on its own testing.
• Against traditional search engines, it performs about 15% better on relevance in those tests.
• Against other AI platforms, it claims about 10% higher response quality.

• For simple questions, the average response time is about 1.2 seconds.
• For complex or multi-part queries, the average response time is about 2.5 seconds.

Accuracy figures are vendor-reported. Treat them as directional, not independently benchmarked. The point that holds up is the source-linked format: every answer ships with citations you can click and check.

9. Team Size, Valuation, and Revenue

Despite its global reach, Perplexity AI runs with a lean, efficient team.

• As of mid-2025, the company had a team of about 38 employees, a sign of its focus on high-impact work.
• It planned to grow the team to about 60 members by the end of 2025 in response to rising demand.
• In January 2024, Perplexity was valued at $520 million.
• By May 2025 the valuation reached about $18 billion, and a $200 million round in September 2025 pushed it to roughly $20 billion, backed by investors such as Nvidia, SoftBank Vision Fund 2, and Jeff Bezos.
• Annual recurring revenue climbed from about $100 million toward roughly $200 million ARR by late 2025, a sign of commercial traction and enterprise demand.

These figures show how small, focused teams, when backed by strong AI and execution, can scale into a multi-billion-dollar business without a massive headcount.

10. The Comet Browser

Comet launched free worldwide in October 2025. It started as a paid beta for $200-per-month subscribers, but now anyone can download it. A free iOS version followed in March 2026 and reached number three on the US App Store within 48 hours.

What Comet Does

• 800+ app integrations for editing docs and analyzing data in the browser
• WebML API support, which runs AI on your device hardware for speed
• A three-tier privacy model
• An email assistant that drafts and schedules

Coming soon: a wider mobile rollout and the Background Assistant, which handles multiple tasks at once.

The waitlist had millions of people before the free release. Download at perplexity.ai/comet for Windows, Mac, and Linux.

Read also: ChatGPT vs Perplexity: which AI assistant is better for coding? (2026 edition)

What the 2026 Model Lineup and Benchmarks Say

The biggest change since the early stats is the model menu. Perplexity is no longer tied to one model. In 2026 Pro and Max users pick from a curated set of frontier models, and a new Model Council compares them side by side.

• Model Council (launched February 2026): ask one question and see answers from several models at once, then keep the best.
• Current frontier models in the picker: GPT-5.2, Claude 4.6 (Opus, Sonnet, and Haiku), Gemini 3.1 Pro, and Grok 4.
• Perplexity’s own models: Sonar, built on Llama 3.3 for fast cited search, and R1 1776, a reasoning model based on DeepSeek R1.
• Open-weight option: Nemotron 3 Super, Nvidia’s open reasoning model, for cost-sensitive workloads.

The takeaway for buyers: Perplexity is now a router over the best models plus live web search, not a single-model chatbot. That is the feature competitors find hardest to copy, because it turns model choice into a per-query decision instead of a lock-in.

When to Use Perplexity vs ChatGPT vs Gemini

No single tool wins every task. Here is a simple decision matrix based on Index.dev testing across research, coding, and team workflows.

Most teams should not standardize on one assistant. Pair Perplexity for research and verification with ChatGPT or Gemini for generation, and let the task decide.

Where Perplexity Wins and Loses in Production

Where it wins

• Technical research and vetting: finding official docs, GitHub examples, and migration guides faster than a Google link list.
• Competitive and market research: Spaces keep sources organized per project, with citations attached.
• Regulated or claim-heavy work: every answer ships with sources, so reviewers can audit the trail.
• Mobile and emerging markets: the Telkomsel bundle and WhatsApp access put cited search in front of users with no separate app.

Where it loses

• Heavy code generation: ChatGPT still produces and iterates on code more reliably.
• Long creative threads: conversational memory across a long session is weaker than ChatGPT.
• Vendor-reported accuracy: the 95% figure is self-reported, so treat it as directional, not a benchmark.
• Free-tier limits: Gemini’s free tier allows more queries for casual users.

Perplexity Features: What Makes People Switch

The stats show growth, but features explain why. Here is what actually works.

Pro Search

• Searches 300+ sources per query vs about 10 on quick search
• Takes 15 to 20 seconds vs 2 to 3 seconds for quick search
• Asks clarifying questions before searching, called copilot mode
• You get 300 Pro searches per day on the paid plan

Example: search "analyze microservices architecture for real-time analytics" and Pro Search returns structured comparisons with links to vendor whitepapers. Quick search just lists pages.

Spaces

• Create separate collections for different projects
• Upload reference PDFs that stay available to the whole Space
• Invite team members with different permission levels
• Switch between personal, link-shared, or team-only privacy

We use Spaces at Index.dev for things like "Competitor Research" and "Developer Hiring Trends." It keeps everything organized instead of scattered across random searches.

Comet Browser

• Launched as a paid beta, went free worldwide in October 2025, free on iOS in March 2026
• Has 800+ app integrations (Google Workspace, Notion, Linear, GitHub)
• Runs AI locally using your device hardware (WebML API)
• Three privacy tiers: public search, private local processing, or hybrid

The big difference from ChatGPT extensions: Comet is a full browser with deep OS integration, not just a plugin.

Multi-Model Access

• GPT-5.2: strong general knowledge and writing
• Claude 4.6 Sonnet: reasoning and code analysis
• Gemini 3.1 Pro: long context and fast responses
• Grok 4: real-time info and current events
• Sonar: Perplexity’s own model for fast cited search
• R1 1776: cost-effective reasoning for high volume

The "Best" mode picks the right model automatically based on your query type, and Model Council lets you compare a few at once.

WhatsApp Integration

• Ask Perplexity through WhatsApp messages
• Get answers with citations sent back as messages
• Popular in Indonesia through the Telkomsel partnership
• No separate app needed

This is useful when you are mobile and do not want to open another app.

File Analysis

• PDFs: extract text, answer questions, summarize
• CSVs: analyze data and spot patterns
• Images: describe content and extract text (OCR)
• Max file size: 25MB on Pro

Example: upload a research paper PDF and ask "What methodology did they use?" Perplexity finds the relevant section and explains it.

How Perplexity Compares to Alternatives

vs Google Search

• Perplexity gives synthesized answers with sources listed
• Google gives a page of links to click through
• Perplexity has no ads or SEO spam
• Google has better local business info and maps

vs ChatGPT

• Perplexity searches the web in real time
• ChatGPT has more limited web access
• Perplexity always cites sources
• ChatGPT usually does not cite where info came from
• ChatGPT has better conversational memory for long threads
• ChatGPT is better at generating code

vs Gemini

• Perplexity lets you use multiple models (GPT, Claude, Gemini all in one place)
• Perplexity has better citation quality
• Gemini integrates natively with Google Workspace
• Gemini’s free tier allows more queries

For Developers (based on Index.dev testing)

• Code documentation: Perplexity finds official docs and GitHub examples faster
• API research: cites specific versions and migration guides
• Tech comparisons: gives multiple source perspectives
• Debugging: searches Stack Overflow, GitHub issues, and docs at once

When developers at Index.dev need quick research for hiring or vetting candidates, we use Perplexity for citation quality. For actual code generation, ChatGPT still leads.

What This Means for Hiring AI Engineers

Perplexity’s growth curve is a hiring signal. Running 1.2 to 1.5 billion cited queries a month needs engineers who can build low-latency retrieval, evaluate models, and keep a citation pipeline accurate at scale. Those skills are exactly what teams now compete for.

The lesson from a 38-person team reaching a $20 billion valuation is that a small group of senior engineers, paired with the right models, can outbuild a much larger one. The constraint is talent, not headcount.

If you are building AI products and need that level of engineer, the practical move is to hire from a vetted senior pool rather than screen thousands of applicants yourself.

For Developers

Index.dev connects the top 1% of senior engineers, selected from a pool of 2.5 million professionals through a five-stage human-led vetting process, with a community of 30,000+ vetted engineers from LATAM and CEE. Roughly a 1.2% acceptance rate, so only the strongest get in.

If you build AI products and want roles that value your skills, join Index.dev and get matched with companies that move fast.

For Clients

Index.dev is an AI-first engineering talent platform that connects companies with the top 1% of senior engineers from LATAM and CEE, around 30,000 vetted developers from a pool of 2.5M+, with matches in 48 hours. Clients have saved $80M+ in hiring costs and see 40 to 60% cost savings on engineering projects, and 97% return for a second engagement.

Hire vetted AI engineers through Index.dev and start building in days, not months.

Final Thoughts

By mid-2026, Perplexity is estimated to process 1.2 to 1.5 billion search queries per month, driven by Pro Search, Comet usage, and mobile growth.

For developers building AI products, Perplexity’s citation-backed research beats Google’s link lists when you need to verify technical claims or compare sources. For code generation, ChatGPT still leads.

The Indonesian expansion through Telkomsel shows Perplexity is thinking beyond English-speaking users. Bundled pricing at about $2.30 per month makes Pro accessible to far more people.

Try Perplexity Pro free or check out the Comet browser to see if the stats match your workflow.