Sacra Logo
View PDF
View Model
San Francisco, CA
Aravind Srinivas
Listed In
Home  >  Companies  >  Perplexity
Perplexity is an AI-native web search startup.












Sacra estimates that Perplexity is at $20M ARR as of April 2024, up 82% from $11M in February 2024 and up 186% from $7M ARR at the end of 2023.

Compare to OpenAI at $2B ARR at the end of 2023, up 900% year-over-year, and Anthropic at $316M annual recurring revenue (ARR) at the end of March 2024, up 1,341% year-over-year from $22M ARR in March 2023.



About 94% of all searches on Google today are 1-5 word queries, both basic information requests and high-intent searches like “pizza near me” against which Google has built a $260B/year business selling ads.

OpenAI's ChatGPT launched with the promise of replacing Google for more complex queries, but with training data that’s still stuck in January 2022 for free accounts, it can’t give up-to-date information.

Perplexity (2022) found initial product-market fit answering complex queries like “How do I provide a California employee with health insurance?” by pulling up the top search results on Bing and passing them into OpenAI’s GPT-3.5 to synthesize the content into an AI-generated summary with citations.

Perplexity has now positioned itself as the “Google Search for knowledge work”—designed for the 0.2% of queries longer than 10 words—monetizing with a $20/mo paid subscription that offers unlimited Copilot searches and usage of the latest models like Claude 3 Opus.

By using GPT-3.5 and Bing, Perplexity was able to quickly launch a compelling search product without needing to build their own web crawler or search index. In 2023, Perplexity began transitioning to open source large language models, fine-tuning versions of Mistral's 7B parameter model and Meta's LLaMa-2 to power its search summaries.

Perplexity has also built out its own search index, with the goal of ultimately replacing its reliance on third-party search APIs.

Business Model


Perplexity operates on a freemium model, offering both a free tier for casual users and a $20/month Perplexity Pro subscription for power users.

The free tier allows users to ask a limited number of questions per month and access a subset of Perplexity's AI models. This tier serves as a customer acquisition funnel, allowing potential subscribers to experience the value of the product before committing to a paid plan.

Perplexity Pro subscribers get unlimited queries, access to Perplexity's full suite of language models (including GPT-4 and Claude), and premium features like the ability to upload files for the AI to analyze and summarize. At $20/month, the Pro plan is priced competitively with other AI subscriptions like ChatGPT Plus.

Perplexity's margins come from the difference between the subscription revenue and the costs of running the language models and search infrastructure. These costs include GPU compute for running inferences on the large language models, storage and compute for maintaining the search index, bandwidth for serving results, and the human costs of the engineering team building the product.

By building on top of open source language models and its own search index, Perplexity can keep its serving costs low compared to paying API fees to OpenAI or search engines. As it scales, Perplexity benefits from the inherent operating leverage in software businesses, where incremental users have near-zero marginal costs.

Longer-term, Perplexity has an expansion motion in providing enterprise versions of its product with additional security and compliance features, priced at a higher tier. The company is also experimenting with new product lines like a browser extension that allows Perplexity to be used across the web for any research task.


The market for AI-powered search is heating up, with several key players, including tech giants like Google and Microsoft, as well as AI startups like Anthropic and OpenAI each vying to dominate the next generation of search.

OpenAI and Anthropic

ChatGPT and Perplexity are battling over who can aggregate consumer demand for AI chat—ChatGPT had a head start, but churn spiked into the 20% range last summer, while Perplexity’s niche approach predicated on weekly usage has resulted in stronger retention dynamics.

That said, OpenAI is now developing its own search index and web crawler, positioning it to build its own comprehensive AI-native search engine—equipped with OpenAI's brand awareness, funding, and technical expertise behind it.

To combat this threat, Perplexity has been investing heavily in fine-tuning open source models like Meta's LLaMa to have its own proprietary model options. Their bet is that owning the search data and user relationships will prove more defensible than the language models themselves.

Anthropic, the startup behind the Claude AI assistant, is also reportedly building its own search product powered by proprietary models and search index. With over $1B in funding and a reputation for AI safety, Anthropic could prove a strong competitor, particularly for privacy-conscious users.

Big Cloud

The biggest long-term competition for Perplexity is likely to be Google and Microsoft, the two dominant players in traditional search. With far greater revenue scale, these giants have the resources to invest heavily in AI to overhaul their search products.

Google already has a chatbot, Gemini, the search data flywheel from their 8.5B searches a day and 1.5B MAUs and ownership of platforms Search, Chrome, and Android—but between their advertiser-focused incentives and big org politics, they’ve barely scratched the surface of building an AI-based consumer product.

Microsoft, meanwhile, has integrated OpenAI's GPT technology into its Bing search engine and is leveraging its close partnership with OpenAI to improve the product.

TAM Expansion

Perplexity’s plan to build the single best user experience for search starts with capturing a slice of Google's most complex queries, then leveraging that proprietary user engagement data to build a data flywheel that continually improves their results.

While Perplexity is built on top of Google search results today, their goal is to use their own data to back into iteratively rebuilding the search stack with AI-native components, from crawler to index to relevance to rankings to personalization.

Looking forward, the key dynamics in understanding Perplexity's growth potential hinges on (1) the shift from traditional search to AI-powered search, and (2) Perplexity's ability to build a differentiated search product.

Search is shifting

The rapid advancements in large language models and generative AI are fundamentally reshaping the search landscape. Users are increasingly expecting search engines to provide direct, conversational answers to their queries rather than a list of links. This shift plays to the strengths of AI-native search products like Perplexity.

Moreover, the rise of AI is expanding the total addressable market for search. While traditional search is primarily used for navigational and informational queries, AI search products are increasingly being used for complex research, analysis, and decision-making tasks. This opens up new use cases and user segments for search products.

As more users adopt AI-powered search assistants for their daily information needs, the market is poised for significant growth. If Perplexity can capture even a small percentage of this expanding market, it could translate into substantial revenue growth.

Perplexity's ability to sustain its growth trajectory will depend on its ability to build a search product that is meaningfully better than the alternatives. The company's current focus on factual accuracy, up-to-date information, and user experience provides a strong foundation.

By continuously iterating on its product based on user feedback and engagement data, Perplexity can create a virtuous cycle where its search results get better over time. As users recognize the superior quality of Perplexity's answers, it could drive increased adoption and retention.

Perplexity's data flywheel strategy - using proprietary engagement data to improve its search stack - could be a significant competitive advantage. As Perplexity builds out its own crawler, index, and ranking algorithms tailored to its users' needs, it could create a search product that is difficult for competitors to replicate.

Furthermore, by expanding into adjacent product lines like browser extensions and enterprise search, Perplexity can increase its touch points with users and capture a greater share of the overall search market. If executed well, these product expansions could significantly increase Perplexity's total addressable market.


This report is for information purposes only and is not to be used or considered as an offer or the solicitation of an offer to sell or to buy or subscribe for securities or other financial instruments. Nothing in this report constitutes investment, legal, accounting or tax advice or a representation that any investment or strategy is suitable or appropriate to your individual circumstances or otherwise constitutes a personal trade recommendation to you.

This research report has been prepared solely by Sacra and should not be considered a product of any person or entity that makes such report available, if any.

Information and opinions presented in the sections of the report were obtained or derived from sources Sacra believes are reliable, but Sacra makes no representation as to their accuracy or completeness. Past performance should not be taken as an indication or guarantee of future performance, and no representation or warranty, express or implied, is made regarding future performance. Information, opinions and estimates contained in this report reflect a determination at its original date of publication by Sacra and are subject to change without notice.

Sacra accepts no liability for loss arising from the use of the material presented in this report, except that this exclusion of liability does not apply to the extent that liability arises under specific statutes or regulations applicable to Sacra. Sacra may have issued, and may in the future issue, other reports that are inconsistent with, and reach different conclusions from, the information presented in this report. Those reports reflect different assumptions, views and analytical methods of the analysts who prepared them and Sacra is under no obligation to ensure that such other reports are brought to the attention of any recipient of this report.

All rights reserved. All material presented in this report, unless specifically indicated otherwise is under copyright to Sacra. Sacra reserves any and all intellectual property rights in the report. All trademarks, service marks and logos used in this report are trademarks or service marks or registered trademarks or service marks of Sacra. Any modification, copying, displaying, distributing, transmitting, publishing, licensing, creating derivative works from, or selling any report is strictly prohibited. None of the material, nor its content, nor any copy of it, may be altered in any way, transmitted to, copied or distributed to any other party, without the prior express written permission of Sacra. Any unauthorized duplication, redistribution or disclosure of this report will result in prosecution.