Most of us can feel the shift happening in real time. Search has stopped behaving like search, and the old habits that defined the past two decades are dissolving faster than anyone in the industry anticipated. Perplexity AI, an answer engine that did not exist three years ago, now handles 1.4 billion monthly queries, a figure that would have sounded implausible even in 2023. It reflects a decisive break from a world where users typed keywords into Google and skimmed sponsored links, to one where they ask a natural-language question and receive a fully synthesised, citation-backed answer in seconds. The disruption is exhilarating and slightly unsettling. Search is changing shape, and none of the incumbents look untouchable anymore.
In this deep dive we explore how Perplexity rose so quickly, why professionals are adopting it at record speed and what the platform’s data tells us about the future of AI-powered discovery. We also examine the risks, the competitive landscape, the underlying technology and the implications for businesses that depend on visibility in a rapidly evolving search ecosystem. By the end, it becomes clear that Perplexity is not a temporary novelty. It is a structural shift in how information is accessed, verified and understood, and its trajectory points towards a very different 2026.
What is Perplexity AI?
Perplexity AI is an AI-powered answer engine designed to deliver accurate, concise and citation-backed responses by synthesising real-time web content through large language models. Unlike generative chatbots that produce unreferenced prose, Perplexity grounds its answers in verifiable sources, typically citing 5 to 10 links across its summaries. CEO Aravind Srinivas has described the platform as “a tool that democratizes access to knowledge” (Aravind Srinivas, Perplexity, 2024), and users often highlight its transparency as the defining characteristic. With support for 46 languages, multimodal query handling, file uploads and image interpretation, the platform serves as a hybrid between a search engine and an AI research companion. This orientation towards factual clarity has given Perplexity an identity distinct from both Google and conversational AI competitors.
Why was Perplexity created in 2022?
Perplexity was born out of a moment of tension in 2022 when traditional search engines were struggling with information overload and early chatbots were producing uncited, sometimes unreliable answers. Srinivas, alongside co-founders Denis Yarats, Johnny Ho and Andy Konwinski, saw a gap emerging: users wanted speed, but they also wanted proof. After witnessing how early LLMs hallucinated without guardrails, the team envisioned a system that would always reveal its working. The August 2022 beta was modest, attracting 3,000 daily queries, but it validated their belief that people craved trustable synthesis during the chaotic post-ChatGPT surge. The system focused on citations from day one, and that philosophical stance has remained central as the platform scaled into one of the fastest-growing AI products of the decade.
How does Perplexity’s real-time answer engine work?
Perplexity’s real-time engine crawls the web for current information, allowing it to generate responses that reflect the latest data, prices and reporting. Unlike static LLMs trained on snapshots, Perplexity continuously retrieves fresh material from news sites, financial feeds and academic archives. Its in-house Sonar model ranks and compresses this information, producing a stitched narrative grounded in citations. The system’s ability to surface recent updates—such as emerging policies or outages—gives it an edge in queries where timeliness matters. This mechanism underpins its reputation for reliability, and it is why users increasingly prefer Perplexity for factual lookups where the difference between outdated and current information can be crucial.
How does a Perplexity query compare with Google and ChatGPT?
Using Perplexity feels different from both Google’s results pages and ChatGPT’s generative output. A typical query begins with natural language—“Compare 2025 EV subsidies”—and within 5 seconds, Perplexity returns a structured, cited response rather than a list of blue links. Google still requires manual navigation, and ChatGPT, although fluent, does not natively cite sources unless prompted. Perplexity maintains focus and avoids drifting into creativity. Its threaded interface allows a sequence of follow-up questions, enabling deeper, contextual research. Users average 12 to 22 minutes per session, suggesting they treat it as a working tool rather than a curiosity. Meanwhile, 90 percent retention within 30 days shows that once people become accustomed to the clarity and speed, they tend not to go back to traditional search experiences. In daily use, 19 percent of its user base relies on Perplexity as a primary search tool, a figure driven partly by growing frustrations with ad-heavy Google pages.
What is Perplexity’s Deep Research mode?
Deep Research is Perplexity’s long-form analysis capability, producing detailed reports across 20 to 30 minutes, combining charts, citations and multi-source synthesis. It operates like an automated researcher, scanning dozens of articles, whitepapers and datasets before condensing them into a coherent document. Professionals use it to analyse markets, industries and competitors at scale, and it has become particularly useful in academic and corporate environments where breadth and accuracy matter. The system’s ability to run in the background, returning a complete digest, has transformed workflows for analysts who once spent hours compiling similar reports manually. It is slower than standard Perplexity queries but dramatically more thorough, and its accuracy benchmarks outperform most one-shot LLM outputs.
Why does Perplexity’s multi-LLM routing matter?
Perplexity is not tied to a single model. Instead, it routes queries to whichever system is best suited for the task, choosing between GPT-5, Claude Sonnet 4, Gemini 2.5 Pro, Grok 4 and its own in-house model Sonar. This approach yields 34 percent higher accuracy in deep-research evaluations than single-model competitors, according to AIMultiple testing. Where ChatGPT is confined to its own architecture and Gemini is locked into a broader Google ecosystem, Perplexity acts as a neutral aggregator. This flexibility has become a key differentiator, especially for experienced users who understand that different models have distinct strengths. For example, Claude is favoured for ethical reasoning, GPT-5 for general intelligence and Gemini for multimodal interpretation. Perplexity orchestrates all of them, making it uniquely adaptable to diverse professional workflows.
Where does Perplexity sit in the 2025 AI landscape?
In 2025 Perplexity occupies a hybrid position that challenges both the search and AI assistant markets. While Google still commands 90 percent of classic search share, Perplexity’s ascent to 1.4 billion monthly queries, up 65 percent year-on-year, represents the first credible crack in the long-standing monopoly. Its valuation of $18–20 billion, alongside 120 million total users globally, signals substantial investor belief in its long-term trajectory. Perplexity is not trying to replace ChatGPT as a creativity engine nor dethrone Google by replicating its ads-driven model. Instead, it is carving out a factual, citation-first niche that appeals to researchers, analysts and corporate teams who need both speed and verification. Analysts at Sacra have labelled it “the first viable Google alternative,” positioning it as a bridge between the old link economy and a new AI-mediated search paradigm.
Which professions rely on Perplexity the most?
Perplexity has become indispensable in fields where accuracy and verifiability outweigh creative expression. Academic researchers report 40 percent adoption, using it to synthesise literature and verify information across sources. Journalists praise its 95 percent citation accuracy, which reduces risk when writing under deadline pressure. Financial analysts rely on its real-time market scanning and automated breakdowns, particularly during earnings seasons when seconds matter. Marketers and SEO practitioners use Perplexity for keyword gap identification, competitive audits and rapid content validation, noting that it outperforms ChatGPT in verifiable outputs by 25 percent for these tasks. Legal professionals use it to scan and summarise contracts, speeding up workloads that once consumed entire afternoons. Even job hunters have turned to Perplexity for personalised opportunity scouting, leveraging its live data feeds to identify openings with greater precision. It is no longer simply an AI tool—it is becoming a professional standard.
How fast is Perplexity growing from 2022 to 2025?
Perplexity’s growth is among the fastest in AI history. In 2025 the platform reported 22–30 million monthly active users, supported by 60–70 million daily queries, far removed from its 3,000-query beginnings in 2022. Monthly visits stand at 153 million, representing 191 percent year-on-year growth, while mobile usage hit 63.53 percent of traffic thanks to the Comet browser extension and mobile-first updates. Peaks of 2.7 million daily visits in August 2024 underscored the growing shift in behaviour, with a stable 2 million+ daily baseline throughout 2025. App downloads reached 50 million, placing Perplexity behind only ChatGPT and Gemini in global AI tool rankings. Its rapid adoption has been fuelled by a mix of professional migration, viral product updates and a growing cultural preference for clean, ad-free interfaces.
What milestones accelerated Perplexity’s rise?
Several pivotal moments between 2023 and 2025 accelerated Perplexity’s adoption curve. A major funding round in December 2024 raised $500 million and pushed the company’s valuation to $9 billion, contributing to a total of $915 million raised by mid-2025. This influx of capital fuelled rapid feature development and infrastructure scaling. The launch of the Comet browser extension in May 2025 proved transformative, driving a 30 percent increase in monthly active users and embedding Perplexity directly into everyday browsing. Another major spike came through the Telkomsel partnership in Indonesia, which increased regional query volume by 20 percent almost overnight. In June 2025, Srinivas publicly announced that Perplexity had surpassed 780 million monthly queries, reinforcing its status as a credible global competitor. These milestones created a compounding effect by attracting professional users who were increasingly dissatisfied with fragmented traditional search workflows.
How does Perplexity compare with ChatGPT, Claude, Gemini and Grok?
Perplexity’s position in the competitive landscape is defined by its performance in factual accuracy, where it leads with 92 percent correctness on web-grounded queries, according to 2025 Vertu rankings. ChatGPT-4o sits at 85 percent, Claude 4 at 89 percent, Gemini 2 at 88 percent and Grok 4 at 87 percent. Perplexity’s insistence on inline citations reduces hallucinations by 40 percent, a defining advantage in environments where precision matters. In AIMultiple’s deep-research benchmarks, Perplexity’s Sonar model achieved 34 percent precision, edging out Claude’s 32 percent, underscoring its ability to synthesise large volumes of information with clarity. Where Perplexity lags is in creativity, where ChatGPT scores 95 percent compared with Perplexity’s 82 percent. Coding is another area where Claude 4 leads at 72 percent, ahead of Perplexity’s 65 percent performance. Speed is competitive across platforms at 3 to 5 seconds per query, although Perplexity’s Deep Research mode can extend to 10–15 minutes due to the complexity of its analysis. Its higher cost for heavy users, at £20 per month for Pro, also factors into comparisons. Nonetheless, for factual, verifiable tasks, Perplexity consistently outperforms.
Model comparison table
| Model | Factual accuracy | Creativity performance | Coding benchmark | Trust score | Cost effectiveness | Average latency |
| Perplexity (Sonar + multi-LLM) | 92% | Moderate | 65% | 4.8/5 | Medium | 3–5 seconds |
| ChatGPT-4o | 85% | Very high | 70% | 4.5/5 | Medium | 2–4 seconds |
| Claude 4 | 89% | High | 72% | 4.6/5 | High | 4–7 seconds |
| Gemini 2 | 88% | High | 68% | 4.4/5 | High | 2–4 seconds |
| Grok 4 | 87% | Very high | 66% | 4.2/5 | Medium | 3–6 seconds |
What are people saying on social media right now?
Social media platforms have played a significant role in Perplexity’s cultural rise, with creators, analysts and professionals frequently showcasing side-by-side comparisons between Perplexity, Google and ChatGPT. Over the past year, mentions have increased by 65 percent, reflecting a shift from curiosity to expectation. TikTok analysts focus on how Perplexity condenses research tasks into minutes, while X threads often praise its transparency for sourcing. Influencers in finance and academia highlight its speed and reliability, using it to demonstrate live policy updates, investment news and scientific summaries. Meanwhile, some creative professionals express frustration with Perplexity’s more clinical tone compared with the conversational warmth of ChatGPT or the humour of Grok. But the dominant sentiment is clear: for factual queries, Perplexity is becoming the default choice. This trend is reinforced by viral walkthroughs of Deep Research and multi-LLM routing, which consistently earn high engagement from knowledge workers seeking to optimise their workflows.
How does Perplexity make money?
Perplexity operates on a freemium model supplemented by enterprise offerings. Its Pro subscription, priced at £20 per month, unlocks unlimited advanced searches, file analysis, model switching and professional tools. Conversion rates now sit between 15 and 20 percent, a notable increase from 10 percent in 2024, and a major driver behind the company’s rising annual recurring revenue. ARR climbed from $80 million at the end of 2024 to between $100 million and $150 million in 2025, and analysts project revenue to reach $656 million by 2026. While subscriptions account for roughly 80 percent of revenue, a developing enterprise business and experimental advertising formats contribute the remainder. Enterprise contracts, in particular, are poised to expand significantly as Perplexity positions itself as a corporate knowledge engine, offering integrations with internal databases and secure environments for confidential research.
What will Perplexity look like by 2026?
Analysts expect Perplexity to enter 2026 with 50–60 million monthly active users, a doubling of its current base, and 3–5 billion monthly queries, positioning it among the most heavily used AI platforms in the world. This trajectory aligns with broader trends in AI adoption, with forecasts suggesting the global AI market will reach $3 trillion by 2030, driven by increased reliance on intelligent search, automation and enterprise integration. Perplexity itself aims for 16 million daily active users by the end of 2025, scaling to over 3 million daily in 2026 as its global footprint expands. Industry observers believe its citation-first approach will become a defining norm for AI systems in regulated sectors, and its hybrid model—bridging real-time search with powerful language models—is beginning to reshape expectations about what an AI assistant should deliver. As analyst Chris Nguyen observed, “Perplexity represents the first structural challenge to search economics in two decades” (Chris Nguyen, AIMultiple, 2025). By 2026, that challenge may become an established reality.
What risks could slow Perplexity down?
Despite its momentum, Perplexity faces several strategic threats. The most immediate is intensified competition from Google, whose AI Overviews now address a growing share of informational queries. Early studies suggest these overviews may already be capturing 20 percent of the question-led traffic that would otherwise flow to Perplexity, a significant encroachment. Regulatory risks are also mounting. The EU AI Act introduces strict requirements for data sourcing and model transparency, potentially raising operating costs by 15 percent. Search itself is becoming expensive as compute costs rise, with Perplexity spending $0.01–$0.05 per query for processing and retrieval. Another vulnerability is dependency on external LLMs; if OpenAI, Anthropic or Google adjust their pricing or licensing terms, Perplexity could face margin pressure. Finally, the company’s valuation, currently around $18–20 billion, may create investor expectations that become difficult to maintain if user retention drops below 90 percent. The next phase of growth will require careful balancing of innovation, cost control and regulatory compliance.
What moves could push Perplexity into the next tier?
Perplexity’s path to the next tier involves strengthening its enterprise footprint, enhancing its proprietary models and expanding its product ecosystem. Analysts expect Perplexity to introduce enterprise-grade knowledge systems built for internal corporate search, positioning it as a challenger to established enterprise AI providers. Opportunities for major commercial contracts exceeding $100 million are likely as Fortune 500 companies seek trusted AI research tools. A significant upgrade to its in-house Sonar model, potentially evolving into a frontier-scale LLM, would reduce reliance on third-party providers and improve differentiation. Perplexity is also a candidate for strategic acquisitions, with SEO platforms such as Ahrefs or data intelligence firms cited as plausible targets. Partnerships with cloud providers like Amazon or Google Cloud could accelerate its international scaling, helping the platform reach the projected 100 million users by 2028. Each of these moves would strengthen Perplexity’s strategic position during a period of rapid AI convergence.
Expert commentary from Blue Ocean Media
“As businesses rethink their search strategies, Perplexity is quickly becoming a visibility battleground. We are already seeing clients increase leads by 50 percent simply by positioning their brands to be cited in AI search results with our AI search optimisation service. This shift is happening faster than most organisations realise, and those who adapt early will maintain an advantage that compounds over time,” says Geoff Parker, Managing Director of Blue Ocean Media (Geoff Parker, Blue Ocean Media, 2025).
AI to make up 50 percent of B2B queries by 2026
As we look toward 2026, the most consequential shift is the projection that AI-driven platforms will handle 50 percent of all B2B search queries, transforming how companies attract visibility, build authority and capture demand. We believe this is not a distant scenario but an imminent reality, accelerated by tools like Perplexity that compress research into seconds and elevate credible sources. Businesses that optimise for AI search now—structuring content for citation, strengthening entity signals and adapting to multi-LLM environments—will see disproportionate gains. Those that delay may find themselves invisible in a search landscape that no longer resembles the link-driven world of the past. Perplexity is not merely a disruptor; it is a signpost for the future of search, and the companies prepared to act today will be the ones shaping that future.
