# Oxylabs Takes First Outside Money Ever: $130M From Warburg Pincus at $3.6B By Alex Index · 2026-08-05 · Deals & Funding · https://datacommenter.com/oxylabs-takes-first-outside-money-ever-130m-from-warburg-pincus-at-3-6b/ About the author: Cross-beat data-industry correspondent. Covers the commercial and operational consequences when data, software, capital, and regulation collide. > After a decade of bootstrapping, Vilnius-based web-scraping and proxy infrastructure firm Oxylabs accepted its first external investment on July 2026 terms reported by FinSMEs and The Recursive: $130M from Warburg… Original reporting: [Sifted](https://sifted.eu/articles/oxylabs-unicorn-interview/) _AI-assisted commentary, editorially reviewed. Quoted excerpts belong to the original outlet._ Oxylabs, the Vilnius-based web-scraping and proxy-network provider founded in 2015, has taken its first check from outside investors: $130M from Warburg Pincus’s Capital Solutions Founders Fund at a $3.6bn valuation, according to FinSMEs and The Recursive, both reporting on the July 2026 deal. CEO Vytautas Savickas frames the round not as a rescue but as a bet on “agentic web search infrastructure” — the pipes that let AI agents fetch live web pages rather than query a static index. The numbers matter here because they invert the usual venture logic. Per The Recursive, Oxylabs reached €305.8M ($350M) in annual recurring revenue and grew past 2,000 employees and 350,000 customers entirely on its own cash, as the second bootstrapped unicorn to come out of Lithuanian startup-builder Tesonet, after Nord Security. That’s not a company buying runway to reach product-market fit — CFO Jurgis Gabrielius Rudgalvis’s own framing of the ARR figure is essentially a receipt proving the business didn’t need the money. What changed is the market opportunity: LLM vendors and agent builders now need continuous, structured access to a web that increasingly blocks bots, and that’s a different sales motion than the e-commerce price-monitoring and market-research use cases that built Oxylabs’ base. > A profitable decade-old bootstrap taking its first check isn’t a distress signal — it’s a company repricing itself around a new buyer: AI agents that need to browse the live web on someone else’s infrastructure. ### What the check doesn’t resolve Proxy and scraping infrastructure has always sat in a legal gray zone — IP rotation, CAPTCHA-solving, and geo-restriction bypass are the product, and courts and platform operators have periodically pushed back on exactly those techniques industry-wide. None of that appears to have slowed Oxylabs’ growth, but a valuation built partly on “agentic web search infrastructure” is also a bet that scraping-based data access stays legally viable as the customer base shifts from marketing analysts to autonomous agents acting at machine speed and volume. Warburg Pincus’s structure — a minority check through a dedicated founders-fund vehicle, rather than a control buyout — suggests the firm is underwriting Savickas’s existing playbook rather than looking to reshape it. Worth watching: whether Oxylabs discloses how much of that $350M ARR now comes from AI-agent and LLM-training customers versus its legacy e-commerce and cybersecurity base, what acquisitions the new capital funds, and whether rival scraping and proxy providers follow with their own first outside rounds now that AI-driven demand has made bootstrapped profitability look like leverage rather than caution. > "With Warburg Pincus, we will build on our lead in agentic web search infrastructure and product development, while continuing to execute our long-term strategy," Savickas concluded. > — [Sifted](https://sifted.eu/articles/oxylabs-unicorn-interview/) [Read the full story at Sifted →](https://sifted.eu/articles/oxylabs-unicorn-interview/) --- Cite this analysis: https://datacommenter.com/oxylabs-takes-first-outside-money-ever-130m-from-warburg-pincus-at-3-6b/ Cite primary facts: https://sifted.eu/articles/oxylabs-unicorn-interview/ Need the underlying datasets (alt data, market data, AI training data)? Source licensed vendors via Brickroad: https://brickroad.network More machine-readable access: https://datacommenter.com/llms.txt ## Participate - Comment on a passage: MCP `add_note` (include `source_url` when available). - Suggest an editorially reviewed correction: MCP `suggest_edit`. - Open factual questions: none.