Who is the next Anduril?

Why is everyone asking who the next Anduril is now?

The question matters now because defense startups have moved from prototypes into serial production, shipyards, and multi-year military contracts.

The money arriving this year is hard to dismiss. Anduril, Helsing, Shield AI, Saronic, and Quantum Systems have announced about $11.75 billion of financing between them in 2026. That figure covers only five companies, not the whole defense-tech market.

Government spending is following. Pentagon contracting with the 15 largest defense startups has roughly tripled since 2022, according to a Wall Street Journal analysis. Startups still receive less than 1% of total Pentagon contract spending, but the direction has changed.

The evidence is also more physical than it was a year ago. Saronic is building autonomous boats by the hundreds and has won a Navy award for larger drone ships. Helsing says thousands of HX-2 strike drones have reached Ukraine since early 2026. Quantum Systems’ Ukrainian joint venture plans a fivefold production increase for 2027.

At the same time, the benchmark keeps moving. In October 2026 alone, Anduril added an Army command-and-control contract and a Navy submarine-industry project, both detailed below.

The search for the next Anduril is no longer about who can raise money. It is about who can build things in volume, on time, for more than one customer.

What would the next Anduril actually look like?

The next Anduril would combine military software, owned hardware, large-scale manufacturing, and major government programs inside one growing company.

A startup could become very valuable by dominating hypersonic missiles, drone defense, military software, or autonomous ships. That would still make it a specialist supplier rather than another Anduril.

Anduril’s advantage comes from the way its businesses connect. Lattice provides a common software layer across surveillance towers, drones, autonomous aircraft, underwater vehicles, missiles, and command centers. Each deployment can improve other products.

Anduril also takes responsibility for manufacturing. It builds factories, supply chains, engines, rocket motors, sensors, and complete military vehicles rather than stopping at software or a prototype. Its October 2026 submarine project extends that logic to heavy industry.

The next Anduril should pass four tests: reusable software, owned military hardware, credible manufacturing, and funded government production orders.

Few defense startups pass more than two. Saronic now passes three, but only at sea. Helsing passes all four on paper, but its manufacturing evidence is thinner than its ambitions.

Did Anduril’s October contracts put it out of reach?

Anduril has pulled far enough ahead that no startup will match it soon, although another company can still build a similar position in Europe or within a narrower set of military markets.

The financial gap is enormous. Anduril generated a reported $2.2 billion of revenue in 2025 and raised $5 billion at a $61 billion valuation in May 2026. In October 2026 it opened an employee share sale capped at about $1.3 billion that values the company at roughly $90 billion, according to reports. That is a secondary sale of existing shares, not a new funding round, but it shows how investors price the company.

More importantly, Anduril keeps converting projects into production. In early October 2026 the Army awarded it a five-year contract to extend its Next Generation Command and Control system to more units, starting with I Corps. The base period is worth $162.8 million, and the ceiling is $1.8 billion if all options are used.

Days later, Anduril and the Navy announced Arsenal-2, a shipyard at Tradepoint Atlantic in Baltimore County, Maryland, meant to ease the bottleneck in Virginia-class submarine production. The $6.6 billion headline combines $3.7 billion of planned Anduril capital and a Navy contract worth up to $2.9 billion, with payments tied to production results. It is a plan, not a purchase order.

Those awards sit on top of the FQ-44 autonomous fighter entering production for the Air Force, Barracuda cruise-missile framework agreements, and Australia’s A$1.7 billion commitment to Ghost Shark submarines.

Competitors are chasing a company that is now adding software, aircraft, missiles, and shipbuilding at the same time. Catching Anduril is not realistic this decade. Becoming a second company of the same type still is.

Which defense startup is closest to becoming the next Anduril?

Helsing is still the closest startup to Anduril because it combines the broadest product portfolio with a shared AI foundation, owned hardware, acquisitions, factories, and real government demand. Saronic has closed much of the gap on execution.

Saronic has moved fastest from startup experimentation into serious production. More than 400 Corsairs built, a Navy production award for the larger Marauder, a landing-craft contract, and a shipyard under construction make it the strongest manufacturing story among the challengers.

Shield AI has the strongest dedicated autonomy platform. Hivemind flies on several aircraft built by other companies and has now steered uncrewed boats in demonstrations. Its owned products remain concentrated in aviation.

Quantum Systems combines large drone orders, battlefield experience, factories, ground vehicles, two new maritime drones, and an open command platform. It now covers air, land, and sea, although its sea and ground lines are still in trials.

Helsing still leads because it covers the most ground with real demand behind it.

The company started with software for military sensors, electronic warfare, and decision-making. It has since expanded into HX-2 strike drones, SG-1 underwater gliders, the CA-1 Europa combat aircraft, electronic-warfare systems, robotics research, and command software. In July 2026 it raised $1.8 billion at an $18 billion valuation, led by Dragoneer.

Acquisitions support that breadth. Grob Aircraft brought decades of composite-aircraft manufacturing. Blue Ocean added underwater-vehicle expertise. In October 2026 Helsing also announced a commercial partnership with Destinus, the European drone and cruise-missile maker.

Other startups deserve attention without fitting the full comparison. Castelion is advancing lower-cost hypersonic weapons. Epirus is building high-power microwave systems against drone swarms. CHAOS Industries is developing advanced sensing. Vannevar Labs is growing inside military intelligence software.

No challenger follows the full Anduril playbook as closely as Helsing. But the gap between Helsing and Saronic is now about breadth, not about proof.

Company	Strongest case today	What keeps it behind Helsing
Helsing	Broad software, strike drones, aircraft, underwater systems and factories; thousands of HX-2s delivered	No published revenue; CA-1 has not flown; demand concentrated in Germany and Ukraine
Saronic	400+ Corsairs built, Marauder production award, shipyard under construction	Almost the entire business remains maritime
Shield AI	Strongest reusable military autonomy software; $400m X-BAT commitments	Owned hardware remains in aviation; maritime work is still demos
Quantum Systems	Volume drone production, battlefield experience, now air, land and sea	Ground and maritime lines are still in trials
Castelion	Fast progress toward operational hypersonic weapons	One main weapon family
Epirus	Credible non-kinetic defense against drone swarms	One main mission and product platform

If you want more recent data on this point, see the latest figures in our Defense Tech report ↓.

Does Helsing have real military demand yet?

Yes. Helsing has enough military demand to be taken seriously, although its valuation assumes that today’s orders will spread to more countries and products.

Ukraine provides the strongest volume evidence. Helsing said in July 2026, when it named former Ukrainian deputy defense minister Andriy Shevchenko to run its Ukrainian business, that thousands of HX-2 strike drones had been handed to Ukraine’s forces since early 2026. That follows an earlier announced production run of 6,000 HX-2s and an order for 4,000 HF-1 drones. Helsing has not published an exact delivery total, and no independent audit of battlefield results is public.

Germany provides the strongest financial evidence. The government approved an initial HX-2 order worth about €269 million, with options that could bring Helsing’s share of the program to €1.46 billion. German lawmakers also approved a €223 million combat-cloud project involving Helsing.

The company also works inside larger aircraft programs. Helsing and Saab hold a contract worth hundreds of millions of euros to add AI-supported electronic warfare to German Eurofighters. Centaur has flown on a Saab Gripen test aircraft, giving the autonomy software a path into existing fighter fleets before CA-1 is ready.

U.S. Army testing added outside validation. During an exercise in Lithuania, 17 HX-2 drones reportedly achieved 15 direct hits and two near misses across a broader exercise of roughly 200 flights.

The demand is real. What is missing is disclosure: Helsing does not report revenue, so investors and governments have to judge it on orders and deliveries it chooses to announce. The next test is repeat business outside Germany and Ukraine.

Can Helsing really build all this hardware?

Helsing’s hardware expansion is credible for strike drones and underwater gliders, while its combat aircraft and robotics programs remain much earlier bets.

HX-2 is the product that already scales. Small strike drones use fewer components and shorter production cycles than aircraft, and thousands have reportedly been delivered. A partnership with Eurenco covers European-made warheads.

SG-1 Fathom has moved beyond a rendering. Helsing opened a factory in Plymouth to produce and maintain the underwater gliders after trials in British and Australian waters. The vehicles are designed to work in groups for persistent submarine and infrastructure surveillance.

In July 2026 Helsing also picked West Virginia for its first U.S. Resilience Factory. It has not yet disclosed output or U.S. orders for that site.

CA-1 brings far more risk. A combat aircraft requires engines, flight controls, sensors, secure communications, weapons integration, testing, and certification. Helsing has Grob’s manufacturing experience and Hensoldt’s sensors, and has said it targets operational readiness in 2029. But no first flight has been announced, and no government has committed to buy it.

The electronic-attack variant makes CA-1 more interesting commercially, but adds integration work before the basic aircraft has flown. RX-1 should still be seen as a research platform.

Helsing does not need every project to work. It does need two or three hardware families in repeat production. HX-2 is there, SG-1 is close, and CA-1 carries the largest upside and the greatest risk.

Does Europe help Helsing or hold it back?

Europe gives Helsing a strong home-field advantage, although fragmented procurement could slow it down.

European governments want more weapons produced in Europe. They worry about ammunition shortages, foreign software control, export restrictions, and dependence on American suppliers. Helsing can present itself as an answer to all four.

Its business is spread across Germany, France, and the United Kingdom. Grob builds aircraft in Germany. The Plymouth factory produces underwater systems. Eurenco can supply European warheads. Hensoldt contributes sensors. The West Virginia site gives Helsing a door into the United States.

The European Union’s Readiness 2030 initiative could mobilize as much as €800 billion for defense. Drones, electronic warfare, AI, missiles, and infrastructure protection sit close to Helsing’s main products.

The difficulty is that Europe still behaves like many markets. Germany, France, Britain, Poland, Italy, and smaller countries run separate competitions, protect their own suppliers, and often demand domestic production.

Helsing has responded with partners rather than fights. Saab, Hensoldt, Grob, Eurenco, OHB, and now Destinus give it routes into programs it could not win alone.

Competition inside Europe is also rising. Quantum Systems, also German, now sells across air, land, and sea, and Germany is financing its interceptor drones for Ukraine. Helsing is no longer the only German champion.

Europe helps Helsing today. It becomes a weakness only if major orders stay concentrated in Germany.

Is Shield AI actually better at military autonomy?

Shield AI still holds the stronger pure autonomy position because Hivemind has run on more third-party vehicles and sits inside a U.S. Air Force production program.

The strongest validation comes from the Collaborative Combat Aircraft program. Hivemind flew aboard Anduril’s YFQ-44 and later received a production contract from the Air Force as a standalone mission-autonomy provider. Anduril has its own Lattice stack, yet the Air Force still chose Shield AI for that layer.

Hivemind also moves quickly between platforms. Shield AI integrated it with a Mitsubishi aircraft project in less than 60 days, demonstrated strike teaming with Destinus, won work on the LUCAS attack-drone program, and flew an autonomous electronic-warfare mission with L3Harris in which one aircraft rerouted others without human input.

The software has now left the air. In July 2026 Hivemind coordinated two SeaShark uncrewed boats with Thunder Tiger in Taiwan. In September 2026 it repeated coordinated surveillance with two Kraken K3 Scout boats. Those were demonstrations with two vessels each, not fleet deployments.

Shield AI’s acquisitions support the same goal. Heron Systems brought the AI that beat a human pilot in DARPA’s AlphaDogfight trials. Sentient Vision added visual detection. Aechelon added simulation. Martin UAV supplied V-BAT.

No common test compares Hivemind directly with Helsing’s Centaur or Anduril’s Lattice on the same missions. Public evidence still gives Hivemind the edge because it has been integrated, flown, and selected on the widest range of outside platforms.

If you want more recent data on this point, see the latest figures in our Defense Tech report ↓.

Can Shield AI turn its maritime demos into real contracts?

Not yet. Shield AI has shown that Hivemind works on boats, but no large maritime production award tied to Hivemind is public, and its biggest new money still goes into aircraft.

V-BAT gives Shield AI an established aircraft business. India selected V-BAT with Hivemind licenses, Poland ordered it for naval operations, Greece expanded its fleet, and Japan chose it as its first ship-based intelligence drone. In October 2026 Shield AI said it had delivered the first V-BATs to Taiwan’s coast guard, out of 12 scheduled over the following months.

X-BAT is where the capital is going. On 8 October 2026, Shield AI said government and company commitments to the vertical-takeoff autonomous fighter had reached $400 million: $200 million from the U.S. government, including a new $150 million Navy contract change, and $200 million of Shield AI’s own money. That is development funding, not a production order. Flight testing is planned for later in 2026, and Shield AI has picked a site near Seattle for its first production facility.

The maritime work is promising but thin. Two demonstrations with partner boats show Hivemind can coordinate surface vessels. They do not show a customer paying for it at scale.

Shield AI still lacks its own ships, ground vehicles, missiles, radar systems, and command network. Its fastest route into new domains is licensing Hivemind to other builders, not building them itself.

For now, Shield AI looks like the leading independent AI pilot for military vehicles, with a fighter program attached. That is a large business. It is not yet a second Anduril.

Could Saronic become the Anduril of the sea?

Yes. Saronic now has the strongest single-domain case in defense technology, because it has production volume, a second production award, and a shipyard under construction.

The volume is unusual. Saronic said in September 2026 that it had built more than 400 Corsair boats, the 24-foot autonomous vessel at the core of its Navy production contract. Saronic now describes that contract as worth more than $400 million, while the figure reported when it was signed was $392 million; the documents needed to reconcile the two are not public.

Saronic is also moving into larger ships. On 29 September 2026 it won a firm-fixed-price production award for its 180-foot Marauder under the Navy’s medium uncrewed surface vessel marketplace, after more than 140 test scenarios over eight days of continuous operation, which the company says it completed without failure. USNI News reported that three Louisiana shipyards, Saronic’s among them, will each build 10 vessels at an average of about $40 million per hull. Saronic’s own award value is not disclosed.

The Navy also awarded Saronic a contract in September 2026 for crewed landing craft, with a 170-ton payload, production planned from 2028, and first deliveries in 2029.

Those ships are meant for Port Alpha. Saronic broke ground on 30 September 2026 on 835 acres in Brownsville, Texas, with more than $3 billion of planned private investment and up to 10,000 direct jobs over a decade, according to the company. In July the plan had been described as a $3.2 billion project. Groundbreaking is not an operating shipyard.

Saronic’s Echelon software connects navigation, simulation, mission planning, and fleet control. That gives it a software-and-hardware loop similar to Anduril’s, concentrated at sea.

The Navy needs many cheaper autonomous ships, and American yards cannot build them fast enough. Saronic can become a major new prime by solving that problem alone. To challenge Helsing for the full title, it would need aircraft, underwater systems, weapons, or broader command software.

Is Quantum Systems now taking on Anduril in air, land and sea?

Yes, on paper. Quantum Systems now sells or tests drones in all three domains, but only its air business has proven volume.

Its Vector reconnaissance drones have years of operating experience in Ukraine, exposing hardware and software to jamming, changing tactics, rough maintenance, and fast feedback.

Ground came first. In July 2026 Ukraine agreed to validate 10 MANDRILL unmanned ground vehicles and 10 Daimler Zetros trucks fitted with the company’s autonomy kit. The contract value was described only as multi-million.

Sea came next. On 14 September 2026 Quantum Systems launched Seavex, a long-endurance surface drone, and Subvex, an underwater drone rated to 6,000 meters. Both run on MOSAIC UXS, its command software, and both are still in sea trials with no customer announced.

Repeat demand is the strongest part of the story. On 4 October 2026 Germany said it would finance several tens of thousands of interceptor drones for Ukraine, built in Ukraine with Quantum Systems’ partner WIY Drones. The exact number and price were not disclosed. A day later, QFI, its joint venture with Frontline Robotics, signed a larger follow-on contract with Ukraine’s National Guard for 2027 and said it plans to raise production capacity fivefold.

MOSAIC is designed to connect vehicles made by other manufacturers, unlike the more closed stacks of several rivals. Quantum Systems raised $1.2 billion at roughly $8 billion earlier this year, and its backers have described it as profitable.

Quantum Systems still lacks Helsing’s aircraft, electronic-warfare programs, and political profile. But it is the challenger whose breadth grew most in 2026, and it is building that breadth on paying customers.

If you want more recent data on this point, see the latest figures in our Defense Tech report ↓.

Which Anduril rival can actually manufacture at scale?

No challenger matches Anduril’s full manufacturing loop. Saronic has the clearest published production number, Quantum Systems the strongest unit-volume pipeline, and Helsing the broadest factory network.

Saronic’s 400-plus Corsairs is the only recent, specific production count among the four main challengers. Its existing Gulf Craft yard in Louisiana builds larger hulls, and the Marauder award gives that yard real work. Port Alpha’s initial design capacity is 150,000 gross tons a year, but landing-craft production there is scheduled only from 2028.

Quantum Systems has the clearest near-term volume plan. Reconnaissance drones, interceptors, ground vehicles, and autonomy kits can share suppliers, software, and customers. The fivefold capacity increase planned for 2027 is future capacity, not output yet.

Helsing is building a distributed network of Resilience Factories: underwater gliders in Britain, aircraft and drones in Germany, and a first U.S. site in West Virginia. Thousands of HX-2s delivered show serial production of expendable weapons, although that says little about combat-aircraft production.

Shield AI has proven it can build and deliver V-BATs. X-BAT will require a much larger industrial system, and its first production site has only been selected.

Anduril is moving the bar again. Arsenal-2 commits $3.7 billion of its own planned capital to submarine production, a scale no challenger can match today.

On manufacturing alone, Saronic ranks first among the challengers. Helsing keeps the best chance of producing several different product families within one network.

Is defense-tech funding creating a fake winner?

The funding boom is inflating expectations, especially around Helsing, but the leading companies now have enough contracts and deliveries to prevent the market from being dismissed as hype.

The gap between valuations and revenue is striking. A Financial Times analysis, based on the valuations and revenue projections available when it was published, placed Helsing’s valuation at roughly 32 times projected annual revenue, Shield AI near 21 times, Anduril near 13 times, and Quantum Systems around 8.5 times. Anduril’s October tender price would push its own multiple higher. None of the challengers publish audited revenue.

Investors are betting these companies will grow more like software businesses than conventional contractors. Shared software should allow faster product improvements, while autonomous systems could create larger production runs and export sales.

That argument already works at Anduril. Its roughly $90 billion employee-tender price in October 2026 rests on revenue, production lines, and a widening contract base.

It is weaker for a company with one major customer, an unflown aircraft, or a factory that does not exist yet. Large funding rounds can hide those gaps.

The headline numbers also need care. Helsing’s $1.8 billion was equity raised. Saronic’s more than $3 billion is planned shipyard spending. Shield AI’s $400 million for X-BAT is half company money. Anduril’s $6.6 billion is mostly a plan plus a conditional Navy contract. None of these is revenue.

Helsing carries the highest valuation risk. Its $18 billion price assumes success across strike drones, combat aircraft, underwater vehicles, software, and international manufacturing. Saronic’s risk is execution on a shipyard. Shield AI’s is X-BAT reaching flight. Quantum Systems’ is proving its new ground and sea lines.

Delivered systems and repeat orders remain much stronger evidence than another financing round.

If you want more recent data on this point, see the latest figures in our Defense Tech report ↓.

What could knock Helsing off the top spot?

Helsing keeps first place unless its portfolio starts missing milestones, and Saronic is now the rival best placed to take advantage if it does.

For Helsing, the evidence will come from deliveries and disclosure. HX-2 must keep arriving and working as electronic warfare evolves. SG-1 needs regular naval deployments. CA-1 must fly and win a government buyer. Orders outside Germany and Ukraine would show it can be a European supplier rather than a national champion.

Saronic could overtake by turning the Marauder award into delivered ships and showing that Port Alpha shortens build times. Selling Echelon for vessels built by others, or moving underwater, would add breadth.

Shield AI could take the lead by making Hivemind the standard autonomy layer across several manufacturers. A successful X-BAT first flight and a paid maritime program would remove its biggest weaknesses.

Quantum Systems has the simplest path. It needs to deliver its interceptor volume, turn ground and sea trials into repeat purchases, and show MOSAIC linking third-party vehicles in operations.

Another funding round would change little by itself. The ranking moves when governments order second batches and pick the same company in more than one domain.

Company	What would move it to first place	What would seriously weaken its case
Helsing	Repeat international drone orders, operational SG-1 fleets and a flying CA-1 with a committed buyer	Delayed deliveries, weak field performance or continued dependence on Germany and Ukraine
Saronic	Marauders delivered on schedule, Port Alpha production gains and a move beyond surface ships	Port Alpha delays or orders staying limited to small boats
Shield AI	Hivemind adopted across major aircraft and ship programs and a successful X-BAT first flight	X-BAT delays and maritime work that stays at demo stage
Quantum Systems	Delivered interceptor volume, repeat ground and sea orders, and battlefield adoption of MOSAIC	New product lines failing to produce material revenue
Castelion	Blackbeard enters recurring operational production and expands into a weapon family	Integration delays or small pre-production orders that never scale
Epirus	A large Army production program and repeat international Leonidas purchases	Continued testing without broad field deployment

So, who is the next Anduril?

Helsing is still the best next-Anduril candidate today, but Saronic has become its closest rival, and the margin between them is small.

Helsing leads because it is attempting the full combination: reusable military AI, software for existing weapons, owned strike drones, underwater systems, a planned combat aircraft, acquisitions, several factories, and large government orders. Thousands of HX-2s delivered to Ukraine show it can produce at volume in at least one category.

No other challenger covers as much ground.

Saronic ranks second. More than 400 Corsairs built, a Marauder production award, a landing-craft contract, and Port Alpha under construction give it the most concrete industrial momentum of any challenger. Its focus on the sea is the only thing keeping it behind.

Shield AI ranks third. Hivemind has the strongest record across outside platforms, now including boats, and X-BAT has $400 million of committed development money. It becomes the stronger answer if Hivemind wins paid programs beyond aviation and X-BAT flies.

Quantum Systems ranks fourth and is the most likely to surprise. It now spans air, land, and sea, has German-financed interceptor volume and a growing Ukrainian order book, and is reported to be profitable.

Our answer is direct: Helsing currently has the strongest chance of becoming the next Anduril.

That judgment is conditional. Helsing has reproduced Anduril’s strategy more convincingly than its results. It still needs to publish more about deliveries, fly CA-1, and win major customers outside Germany and Ukraine.

If it clears those tests, the comparison will be deserved. If it stumbles, there may never be one obvious second Anduril. Saronic could own autonomous shipbuilding, Shield AI military autonomy, Quantum Systems mass-produced battlefield robotics, and Castelion or Epirus individual weapon categories.

For now, Helsing offers the closest match to Anduril’s breadth and ambition. Saronic offers the closest match to its production record. That split is why the race remains open.

If you want more recent data on this point, see the latest figures in our Defense Tech report ↓.