Perplexity AI

Perplexity AI is an American artificial intelligence company, founded in 2022, known for its AI-powered answer engine that searches the web and summarizes results in plain language with linked citations, rather than returning a list of links. Based in San Francisco, the company was started by Aravind Srinivas, Denis Yarats, Johnny Ho and Andy Konwinski, engineers with backgrounds in AI research at places including OpenAI and Google DeepMind. Since its main product launched in December 2022, Perplexity has grown from a small research tool into a company valued at roughly $20 billion, while also drawing lawsuits from major publishers over how it gathers and reuses their content.

History of Perplexity AI

Perplexity AI was founded in August 2022. Its main search product launched that December, combining a large language model with live web search so it could answer questions in natural language and cite its sources, rather than requiring users to click through a list of links the way a traditional search engine does. The company reported about 2 million unique visitors in February 2023 and had grown to roughly 52 employees by 2024, a small headcount relative to its rapidly rising valuation, and usage and investor interest grew quickly from there. Chief executive Aravind Srinivas has described Perplexity’s goal as building an “answer engine” rather than a conventional search engine.

How Perplexity’s AI search works

When a user asks a question, Perplexity searches the live web, reads a set of relevant pages, and uses artificial intelligence to generate a written answer that cites the specific sources it drew from. Paying subscribers can choose which underlying model answers their query, including third-party models from OpenAI, Anthropic and Google, or Perplexity’s own in-house models, Sonar and R1 1776. This approach relies on a technique sometimes called retrieval-augmented generation, where a machine learning model’s answer is grounded in freshly retrieved documents instead of only what it learned during training, which is meant to reduce outdated or invented answers compared with a model working from memory alone.

Perplexity’s products

Beyond its original search engine, Perplexity has expanded into a wider set of tools. Perplexity Pro, a paid subscription, adds access to more models and features such as searching a user’s own uploaded documents alongside the web. In July 2025 the company launched Comet, a web browser built around its AI assistant, which it made free to all users that October. In February 2026 it introduced Perplexity Computer, an AI agent able to carry out multi-step tasks across different underlying models, followed two months later by a version designed to run on a user’s own device. The company has also added finance data, shopping recommendations, and in January 2026 signed a three-year, $750 million cloud computing agreement with Microsoft to expand its computing capacity.

Funding and valuation

Perplexity has raised money from investors including Nvidia, Databricks, and Jeff Bezos. Its valuation climbed from about $520 million in January 2024 to roughly $9 billion by December 2024, then to about $14 billion in mid-2025 and around $20 billion by September 2025, with reports of a further round valuing the company near $21 billion in early 2026. The soccer player Cristiano Ronaldo also took a stake in the company and entered a brand partnership with it in December 2025.

Lawsuits and controversies

As Perplexity grew, several news and reference publishers accused it of using their content without permission or compensation. Forbes, Dow Jones, the New York Post, the New York Times, the BBC, Encyclopaedia Britannica, Merriam-Webster, Reddit, CNN, and multiple Japanese newspapers have all raised objections, and a number of them filed lawsuits alleging copyright infringement or improper scraping of their websites. Perplexity has generally responded that it is aggregating publicly available information and that individual facts cannot be copyrighted, while also launching a revenue-sharing program for some publishers in 2024. Separately, the web infrastructure company Cloudflare published research in August 2025 alleging that Perplexity used undeclared crawlers to access sites that had blocked it, a claim Perplexity disputed.

Perplexity AI vs. traditional search engines

A traditional search engine like Google mainly returns a ranked list of links and leaves the user to read each page and form their own answer. Perplexity instead has an AI system read multiple sources itself and return a single written summary with citations attached, which can save time but depends on the underlying model correctly representing what its sources actually said. This difference is also why Perplexity has faced more direct friction with publishers than a conventional search engine typically does, since its answers can reduce the need for a reader to visit the original site at all.

FAQs

What does Perplexity AI do? Perplexity AI runs an AI-powered search engine that answers questions in plain language by reading web sources in real time and citing where its information came from.

Who founded Perplexity AI? Perplexity AI was founded in 2022 by Aravind Srinivas, Denis Yarats, Johnny Ho and Andy Konwinski, and Srinivas serves as chief executive.

Is Perplexity AI free to use? Yes. Perplexity offers a free tier without registration, alongside a paid Perplexity Pro subscription that unlocks additional models and features.

Why have publishers sued Perplexity AI? Several news and reference publishers, including the New York Times, the BBC, Encyclopaedia Britannica and CNN, have accused Perplexity of using their content without permission, and some have filed copyright lawsuits that remain in progress.

What is the Comet browser? Comet is an AI-powered web browser Perplexity launched in July 2025, built around its assistant so users can ask questions, summarize pages, and complete tasks directly inside the browser; it was limited to top-tier subscribers at first and made free to everyone in October 2025.

Editorially reviewed by Specialty Digest Editorial TeamLast reviewed November 9, 2026Researched and drafted with AI assistanceReport an issue