Pentagon Drone Dominance Program Splits 60,000-Drone Order Across Nine Startups

The Pentagon has split an order for roughly 60,000 low-cost attack drones across nine privately held startups, the first concrete purchasing signal to emerge from the Pentagon Drone Dominance Program. Results from Gauntlet II, the second competitive round in the program, were published this week and name five winners in each of two mission areas: long-range strike and close-in assault. Ten finalists came through August testing at Fort Carson, Colorado. None of them trades on a public exchange.

The program is a $1.1 billion push to buy attritable one-way attack drones by the tens of thousands. Gauntlet II accounts for about $300 million of that total, and the top scorer in each mission area earns production orders. The Pentagon attached one non-negotiable condition to entry: the aircraft could contain no Chinese-made components or materials.

Perennial Autonomy finished first in deep strike with 80.1 points. Hyperscale followed at 70.9, Neros at 69.8, Skycutter at 66.6 and Swarm Defense Technologies at 66.0. In close quarters battle, Neros took the top spot at 88.1, with ORQA US, XTEND Reality, Vector and ModalAI filling the remaining places. Point totals were published for the deep strike board only; the lower close-quarters ranks appeared without scores.

Inside the Pentagon Drone Dominance Program Leaderboards

Neros is the only company on both boards. It won close quarters outright and placed third in deep strike, a combination no other finalist managed. The company already supplies Ukraine with 6,000 of its Archer FPV drones, which gives it a production record that most rivals on these lists cannot match.

Perennial Autonomy arrived by a different route. Eric Schmidt founded the company as White Stork to build drone interceptors for Ukraine, and its Hornet drone now tops the long-range category. The backing is notable, though the placement rests on test scores rather than investors.

Skycutter is the sharpest reversal. The British manufacturer won Gauntlet I in March with 99.3 points. Six months later it placed fourth in deep strike at 66.6 and did not appear on the close-quarters board at all. Scores are not directly comparable across rounds and the requirements shifted between them, so the gap is not proof of decline. It is still a reminder that one competition win carries no lasting advantage.

The two mission areas are not interchangeable. Long-range strike covers engagements out to 15 kilometers, while close quarters battle targets tactical assault in confined environments. Endurance, communications links and warhead sizing differ enough that a strong showing in one category says little about the other, which is part of why only Neros placed on both.

Ukrainian manufacturers reached no top slot despite fielding systems that have been used in combat at scale. These screens reward documented supply chains and reproducible manufacturing, and combat experience satisfies neither on its own.

Gauntlet II was not the first filter. The program ran an initial round in March, and the ten finalists named this week emerged from a field that had already been narrowed before the August trials at Fort Carson. The five-place structure in each category matters for procurement: the Pentagon is not choosing a single airframe, it is building a shortlist it can order against repeatedly.

Nine Vendors and One Fragile Supply Chain

Spreading a 60,000-unit requirement across nine suppliers hedges against any single design failing. It also divides the volume. Each vendor carries the fixed cost of qualifying non-Chinese motors, flight controllers, batteries, optics and radios, and those costs spread over fewer airframes when nine companies share the pool. Expect unit prices to stay high and spare-part logistics to get complicated.

One-way attack drones are designed to be expended, which changes what the buyer optimizes for. Survivability features, maintenance intervals and long service lives count for less than unit price, production rate and the ability to replace losses within weeks. A requirement that parts come from outside China runs against that logic, because the cheapest components on the market are usually the ones being excluded.

I think the leaderboard is the least interesting document the Pentagon Drone Dominance Program has released. The consequential decision was the component-origin rule, which turns a drone contest into an industrial policy instrument. A vendor that cannot source parts outside China must build them or qualify new suppliers from scratch, and neither path is cheap or quick.

The Pentagon is buying a supply chain as much as it is buying airframes. Volume is what justifies the fixed cost of domestic qualification, and domestic qualification is what makes volume survivable in a conflict. The $1.1 billion ceiling and the 60,000-unit target only work as a pair.

The arithmetic deserves a look. Spending the entire $300 million Gauntlet II round on airframes alone would imply a unit cost near $5,000, and the round also funds testing, evaluation and program overhead, so the real hardware ceiling sits below that. That is a demanding price for an aircraft required to fly on non-Chinese parts.

Sustainment is the unresolved question. Nine vendors mean nine sets of spares, ground equipment, firmware and training pipelines, and a battlefield customer cannot absorb that complexity at the scale of 60,000 airframes. How the pool will be managed once production orders are placed has not been spelled out, and that gap is where the cost overruns of the next few years will originate.

The Public-Market Signal

Publicly traded Red Cat and Ondas both missed the cut in their respective mission areas, and all ten finalists are privately held. Vector said it received orders for up to 5,000 drone systems plus its HV-ESAD arming hardware. XTEND said it delivered 120 STRIKER systems for testing. Investors hoping for exposure to this round through listed equities got nothing from it.

The order itself has not been allocated. The 60,000 figure describes the pool the Pentagon intends to buy from rather than a committed quantity for any single vendor, and the production awards tied to the two top scores will set the real demand curve for the nine companies on these boards.

The strongest counter-argument is that rankings are a weak proxy for anything durable. Scores measure performance against one specification at one moment, and Skycutter's slide from first to fourth within two rounds shows how fast they move. That objection is fair, and it cuts against the public-market read as much as against the winners. The defensible conclusion is narrower: this round established a vendor pool, not a market position.

For anyone tracking defense autonomy as an investment theme, the useful lesson is that the qualification gate now sits upstream of the contract. A company can hold a strong product and still be invisible to this buyer if its bill of materials fails the origin test. That makes supplier disclosure and component provenance a diligence item rather than a compliance footnote.

Why this matters

The rankings will be forgotten faster than the procurement architecture they create. The Pentagon Drone Dominance Program now holds a pool of vendors that have cleared a non-Chinese supply-chain screen, and it can scale orders across that pool without reopening the qualification question each time. That template is worth watching well beyond drones, because component-origin rules tend to migrate from one procurement category to the next. The number to track is the unit price the two top scorers lock in when production orders arrive; that figure will show whether domestic drone manufacturing is a business or a subsidy.

Related Articles

• Pentagon's AI Counter-Drone Marketplace Puts Procurement Speed to the Test
• Who Wins the DHS Counter-Drone IDIQ: The New Economics of Drone Defense
• Smack's $61M Series B shows Pentagon AI vendor diversification is paying off

✔Human Verified

Researched and cross-referenced against primary sources by the Bytevyte editorial team. This article was generated with the assistance of artificial intelligence and reviewed by the Bytevyte editorial team.