Perplexity CEO Sets 2028 IPO Target, Betting AI Search Can Wait While Rivals Rush In

Perplexity CEO Aravind Srinivas drew a line in the sand this week: the AI search company is going public in 2028, regardless of how the IPOs of OpenAI and Anthropic play out. “Agnostic of these two companies, we were planning for something in 2028, so that still remains the case,” Srinivas told CNBC’s Arjun Kharpal in an interview that aired June 9, as reported by CNBC.

The commitment transforms what was previously a loose “not before 2028” stance into a concrete target. And it comes at the most crowded moment in AI IPO history. SpaceX goes public this week at a target valuation of $1.75 trillion. Anthropic filed confidentially in late May at roughly $1 trillion. OpenAI filed June 8, expected to target around $1 trillion. Combined, the three listings represent nearly $4 trillion in market value — the largest transfer of private‑to‑public wealth in a generation.

Perplexity’s last private valuation, by contrast, was $20 billion. Srinivas is betting that waiting two more years lets the company build a bigger, healthier business — and arrive at the public markets on its own terms.

Perplexity has the runway to wait. Annualized recurring revenue hit $500 million as of April 2026, up 335% year‑over‑year, according to financial data tracked by.² Monthly active users surpassed 100 million, with over 20,000 enterprise clients. The company’s ARR trajectory tells the story: $63 million in December 2024 to $500 million in April 2026 — an 8x climb in 16 months.

A February 2026 pricing overhaul — adding a usage‑based layer on top of existing Pro ($20/month) and Max ($200/month) subscriptions — doubled revenue in a single quarter, Yahoo Finance reported. The company has also signed a $750 million infrastructure deal with Microsoft Azure, giving it the compute muscle to scale independently.

Dmitry Shevelenko, Perplexity’s chief business officer, told Quartz that holding 2028 as the earliest date has let the company “build a healthy, high‑growth business” rather than optimize for a near‑term IPO window.

Srinivas was notably bullish on the upcoming Anthropic and OpenAI IPOs — despite them being competitors. “I think it’s important for the AI industry that these IPOs go well, and I actually think they will go well, because they’re doing well,” he told.¹ “We are very happy for their success, because their success means they’ll be able to invest more into frontier model development. And every time AI gets better, Perplexity gets better.”

This is Perplexity’s structural advantage: unlike frontier labs that build foundation models, Perplexity is an AI applications company. It ingests and cites from multiple models — GPT‑4, Claude, its own — to deliver real‑time search results. Every advance in the underlying models makes Perplexity’s answer engine more useful. When frontier labs spend billions making AI smarter, Perplexity gets the upgrade for free.

Srinivas did acknowledge risk, telling ¹ there would be “ripple effects if they don’t go well, like there is no sugar coating on that.” He added that SpaceX’s IPO this week “will definitely be a leading indicator of how Anthropic or OpenAI will go out.”

Perplexity occupies a unique niche as the only major independent AI‑native search competitor to Google. While Google has launched its own AI overviews and the its biggest search revamp in 25 years with Gemini, Perplexity’s pure‑play AI answer engine model — providing cited, real‑time answers from multiple sources — has carved out a loyal user base of 100 million monthly actives.

The company’s growth puts it in a different category from the model builders. It raised roughly $1.7 billion across 11 funding rounds from investors including SoftBank, NVIDIA, Jeff Bezos, and Accel. Its last round, in September 2025, valued the company at $20 billion, PYMNTS reported.

Waiting until 2028 is not risk‑free. By then, OpenAI and Anthropic will have had up to two years as public companies, with the capital, talent, and market visibility that brings. Both could expand their own search capabilities — OpenAI’s ChatGPT already has a browse mode, and Anthropic’s Claude recently added web access. The line between “AI search company” and “AI model company with search” is blurring.

Srinivas addressed this obliquely when he told:¹ “If for six months you don’t see a model capability advance from one of these two companies, then it’s a problem for them.” The implicit argument: as long as the frontier labs keep pushing AI forward, Perplexity’s application‑layer advantage holds. The risk is what happens if the frontier labs decide to compete on the application layer too.

For builders watching the AI IPO wave, Perplexity’s 2028 bet is a different kind of signal. While OpenAI and Anthropic are racing to capitalize on peak AI hype, Perplexity is betting that the AI search market will be bigger in two years — and that it’ll be ready to own it.