Horizon3.ai Valuation 2026: How a $650 Million Startup Reached $2 Billion in 14 Months

$2 billion is what investors say Horizon3.ai is worth after its $250 million Series E closed on August 3, 2026 — a tripling of the $650 million price set at its Series D just 14 months earlier, on a business that is still privately held and still not profitable by disclosed measures.

The round is the clearest evidence yet that venture money is chasing a specific bet: that AI-driven offense will outrun traditional, human-run penetration testing, and that whoever owns the automated version of that job captures a market currently measured in the tens of billions. Horizon3.ai's pitch is built on autonomy — its NodeZero platform runs AI agents against a live production network the way a real attacker would, all day, every day, instead of the once-a-year manual pentest most enterprises still budget for. TechCrunch reported the round on the same day Horizon3.ai and its investors announced it.

Horizon3.ai Valuation in August 2026: What the $2 Billion Mark Rests On

Horizon3.ai's Series E priced the company at more than $2 billion, up from the $650 million mark its Series D set in June 2025 — roughly a 3x markup in 14 months, according to Forbes and the company's own press release. The $250 million round was oversubscribed and co-led by existing backers NightDragon and NEA, with new money from Acrew Capital, Sapphire Ventures, Singapore state investor EDBI, and defense contractor SAIC joining returning investors Craft Ventures and Qualcomm Ventures.

Founded in San Francisco in 2019 by CEO Snehal Antani — a former CTO of the Joint Special Operations Command and of Splunk — and co-founder Anthony Pillitiere, Horizon3.ai built its early team out of U.S. Special Operations and national-security veterans. That pedigree is central to the sales pitch: the company markets NodeZero as software built by people who used to break into networks for a living, not a security vendor that added an AI feature after the fact.

The Full Funding History: $2.5M Seed to $2B in Six Years

Every disclosed round Horizon3.ai has closed got larger, and the gap between rounds got shorter. The company went more than three years between its $5 million Series A and its $40 million Series C; it went barely 14 months between its $100 million Series D and its $250 million Series E.

Sources: TechCrunch on the 2023 Series C; TechCrunch on the 2025 Series D raise; TechCrunch and Help Net Security on the Series E; Tracxn for the cumulative total across all eight rounds including undisclosed bridge financings. Series A and B amounts and dates per Crunchbase round records.

The Growth Numbers Behind the Markup

A 3x valuation jump in 14 months needs growth to back it up, and Horizon3.ai's disclosed numbers move in the right direction, even if they are not audited. In its March 2026 update, the company reported 102% year-over-year ARR growth and said NodeZero had by then run more than 225,000 production-safe pentests. By the August Series E announcement, that ARR growth rate had accelerated to 120% year over year, with annual recurring revenue approaching $100 million, the pentest count past 300,000, and the customer base above 7,000 organizations — up from roughly 3,000 at the time of the June 2025 Series D, per SecurityBrief.

Horizon3.ai has also disclosed retention figures that matter more to a SaaS investor than the headline growth rate: 125% net dollar retention and 94% gross dollar retention, according to the company's Series E materials. Net retention above 100% means existing customers are spending more over time — in this case, likely a mix of testing more of their network and expanding from pure pentesting into broader exposure-management use cases — which is what lets a company keep compounding ARR even if new-logo growth slows.

The Market Category Horizon3.ai Is Betting On

Gartner renamed this category in its 2026 Market Guide, folding Breach and Attack Simulation and automated penetration testing into a single segment called Adversarial Exposure Validation, or AEV. That naming shift matters commercially: it is the label enterprise security buyers now search for in RFPs, and Horizon3.ai, Pentera, and xBow are all positioned as AEV vendors rather than as niche pentest-automation tools. The broader exposure management market — the umbrella category AEV sits inside — is estimated at roughly $3.5 billion to $5.1 billion in 2026 depending on the research firm, according to Coherent Market Insights, with the narrower AEV segment alone projected to reach $2.5 billion by 2026 at a 35% compound annual growth rate.

Horizon3.ai's own customer base illustrates why buyers are moving off annual manual pentests: the company said NodeZero now runs continuously across hospitals, manufacturers, financial-services firms, and critical-infrastructure operators, spanning managed service providers that resell it to small businesses up through Fortune 10 enterprises. That range is a double-edged data point for valuation purposes — it shows genuine market breadth, but it also means the reported 7,000-plus customer count blends very different contract sizes, and Horizon3.ai has not disclosed an average contract value or how revenue splits between its smallest and largest accounts.

The Series E also lands inside a much bigger wave of money moving into the category. Cybersecurity startups raised $10.6 billion in venture financing across all stages in the first half of 2026, and AI-native security specifically overtook identity management as the fastest-growing VC category in Q1 2026, pulling in $4.1 billion at 47% year-over-year growth, according to Crunchbase News. Data-security startup Cyera raised $600 million at a $12 billion valuation that same June, which puts Horizon3.ai's $2 billion mark toward the middle of a pack of AI-security rounds that have gotten dramatically larger across 2026, not at the extreme high end.

Horizon3.ai vs Pentera vs xBow: How the Autonomous Pentesting Market Is Priced

Horizon3.ai is not alone in this category, and the two clearest comparisons show how differently the market is pricing similar businesses. Israeli rival Pentera — the older company, formerly known as Pcysys — has raised roughly $250 million in total and was valued above $1 billion after a $60 million round in March 2025, with ARR that PitchBook and industry reporting put above $100 million as of January 2026. xBow, a Seattle-based startup founded by Oege de Moor — previously known for creating GitHub Copilot and GitHub Advanced Security — hit unicorn status on March 18, 2026 with a $120 million round led by DFJ Growth and Northzone at a valuation Bloomberg reported as above $1 billion, then added a $35 million extension in May 2026 from investors including Accenture Ventures and Nvidia's NVentures.

Horizon3.ai carries roughly double Pentera's and xBow's valuation on similar total capital raised, a gap its ARR-near-$100M figure only partly explains since Pentera reports comparable revenue on a smaller check.

What the headline misses

A 3x markup in 14 months on a company with no disclosed profitability is not, by itself, proof of anything except that investors believe the category will keep growing. Horizon3.ai has not published audited financials, and every growth figure in this piece — ARR, retention, customer count, pentest volume — comes from the company's own announcements, not from an independent filing. Startups raising at accelerating valuations while burning cash to fund international expansion is a familiar 2021-style pattern, and the fact that it is happening again in 2026, in a market that has otherwise gotten more disciplined about growth-at-all-costs, is worth flagging rather than waving through.

There is also a competitive risk the round doesn't resolve: Pentera has comparable revenue on roughly half the capital raised, and xBow is newer, cheaper, and founded by someone with a track record of building developer tools that reached mass adoption. If any of the three shows it can win enterprise deals against the other two at a materially lower cost, the $2 billion mark on Horizon3.ai — the highest of the group by a wide margin — is the one with the most room to compress. One read on this: Horizon3.ai's premium may reflect its Special Operations-heavy sales narrative and its bigger customer base as much as any structural product advantage, and that kind of premium is usually the first thing a downround corrects.

What Would Move the Number From Here

Three things would test the $2 billion mark. The first is whether ARR growth holds near 120% through the next two quarters or decelerates the way most SaaS growth curves eventually do — Horizon3.ai's own numbers show acceleration from 102% to 120% between March and August 2026, an unusual pattern that a buyer should expect to normalize. The second is customer concentration: a company that serves both managed service providers reselling to small businesses and Fortune 10 enterprises has a very different renewal risk profile depending on which segment is actually driving the ARR, and Horizon3.ai has not broken that out. The third is competitive response — SAIC, a defense contractor, is now both an investor and a potential channel partner, which could accelerate Horizon3.ai's push into government and critical-infrastructure accounts where Pentera and xBow have less of a foothold.

Watch for a Series F or a strategic acquisition announcement as the next signal. Rounds this size, this close together, in a two-company-plus-one-upstart market usually resolve one of two ways: one player pulls decisively ahead on revenue and the others get acquired, or a larger cybersecurity platform vendor buys its way into the category outright the way Docker, Cisco, and CrowdStrike have all done with smaller point-solution acquisitions in adjacent categories over the past two years.

The SAIC relationship is the detail most likely to matter first. SAIC is a government IT and defense contractor with $7.48 billion in fiscal 2025 revenue, not a typical venture fund, and a strategic check from a company that already sells into the federal agencies Horizon3.ai wants as customers functions less like capital and more like a distribution deal with an equity kicker attached. If SAIC starts reselling or bundling NodeZero into its own federal contracts within the next few quarters, that would validate the $2 billion mark faster than any ARR update could — and if it doesn't, the investment will look more like a defensive hedge by a contractor that didn't want a rival to lock up the relationship first.

The Bottom Line

Horizon3.ai enters September 2026 with $2 billion-plus behind it, ARR approaching $100 million, and a growth rate that is still accelerating rather than slowing. That combination is genuinely rare and explains why the Series E was oversubscribed. It does not, on its own, explain why Horizon3.ai trades at roughly double the valuation of two competitors reporting similar revenue on similar capital raised — a gap that a Series F, an IPO filing, or a competitor's breakout quarter will eventually have to justify or close.