Perplexity AI Pre-IPO: From Answers to Autonomous Workflows

From Answers to Autonomous Workflows

Perplexity AI launched in 2022 with a precise thesis: search should return answers, not links. Instead of ten blue links and a wall of text to sift through, users get a synthesized response drawn from live data, with citations. That product now handles roughly 780 million queries a month. The origin story is straightforward. What matters now is the structural shift happening underneath it.

What Is Perplexity AI

Perplexity was founded in San Francisco by CEO Aravind Srinivas, who previously worked at OpenAI, DeepMind, Google Brain, and Berkeley AI Research. The company's core product is an AI-native answer engine that aggregates real-time search data and uses large language models to produce conversational, cited responses to complex multi-sentence queries.

The gap it filled was specific. Traditional search engines were engineered for short keyword queries and optimised around advertiser-driven link hierarchies. Early generation LLMs like GPT-3 were strong on synthesis but cut off from live data, producing confident answers that were often months or years out of date. Perplexity addressed both sides: real-time retrieval, structured reasoning, and source attribution in a single output.

Valuation and Funding

Perplexity raised at a $9 billion valuation in a 2024 funding round, with investors including SoftBank, Jeff Bezos, Nvidia, and IVP. The company is widely expected to pursue a public listing, making it one of the more prominent AI pre-IPO names currently tracked in secondary markets.

Revenue grew 800 percent year-over-year in 2024, driven primarily by subscriptions. The freemium model scales from free limited access through to $200 per month for full access to the company's enterprise and agentic tiers. API access and contextual advertising supplement the subscription base. Margins remain under pressure from compute costs and the proprietary search index infrastructure the company is building in-house.

Perplexity Computer: Autonomous AI Workflows

In early 2026, Perplexity launched Perplexity Computer. The product is not a chatbot extension. It is an autonomous agent layer: a system that takes a high-level goal, decomposes it into subtasks, routes those subtasks across different AI models, and executes the full workflow without human input at each step. Research, coding, document generation, and analysis can all be initiated and completed in the background.

This is a structurally meaningful move. Most AI products currently sit in one of two categories: question-answering interfaces, or single-agent tools that require continuous prompting. Perplexity Computer attempts to build a third lane, closer to a background project coordinator than a conversational product. Whether it executes reliably at enterprise scale remains to be tested. The architectural direction is clear.

Traction and Distribution

The user numbers are real and independently verifiable. Thirty million monthly active users. 780 million queries processed in May 2025 alone. Mozilla Firefox ships Perplexity as a default search option. Samsung's Galaxy AI devices include Perplexity integration. A multi-year licensing agreement with Getty Images signals moves toward licensed content infrastructure rather than scraped data.

On the product side, the company has shipped a standalone browser called Comet, an email assistant, enterprise APIs, and a Samsung TV app. That breadth is either evidence of strong product velocity or a company testing multiple distribution bets simultaneously. Given the funding runway and the team's background, both readings have merit.

Intellectual Property and the Data Moat

Perplexity's defensible asset is not the models. Open-source model foundations are increasingly commoditised across the AI industry. The moat is the query corpus. Every search interaction generates signal: what users ask, what responses they engage with, what follow-up questions they form. That feedback loop trains Perplexity's fine-tuned models and improves its proprietary search index over time. The data flywheel compounds. It is not easily replicated from scratch regardless of capital.

The company is also building its own search index rather than relying entirely on third-party search APIs. That is an expensive and technically complex undertaking, but it reduces dependency on infrastructure Perplexity does not control and improves retrieval quality for the kinds of complex, multi-source queries its users generate.

Why Pre-IPO Access Is Relevant Now

AI-native search is not a speculative category. The product works, the user base is large, and the enterprise workflow play through Perplexity Computer represents a logical and high-value extension of what is already a functioning business. If autonomous AI workflows become the standard mechanism for enterprise research, analysis, and execution, Perplexity enters that stack with a live search layer, a large user base, and a proprietary data advantage already in place.

The pre-IPO window is the relevant consideration. Companies at this stage of traction and valuation trajectory typically close to outside access well before a listing. The current availability on secondary markets reflects a specific and time-limited point in the company's capital structure.

Perplexity pre-IPO access is live on WLTH: https://app.wlth.xyz/pre-ipo-access/tperplexity

WLTH provides tokenised economic exposure to pre-IPO opportunities via 1:1 backed on-chain structures. Access is subject to eligibility. Review full terms before participating.