AST SpaceMobile is moving beyond satellite manufacturing and spectrum
strategy into one of the space industry’s most difficult businesses:
orbital launch. The company plans to build Blacklift, a three-stage
rocket designed around its large BlueBird direct-to-device satellites,
using engines and engineering support bought from Firefly Aerospace.
The development, first identified for this analysis through SpaceNews,
is consequential because it reframes launch access as a strategic asset
for constellation operators. SpaceX already owns both its rockets and
Starlink network. AST is now attempting a narrower version of that
integration, while paying another launch company to accelerate the
hardest propulsion work.
The plan could reduce exposure to crowded launch manifests. It also
asks a satellite operator to become a reusable-launch developer on an
exceptionally compressed timetable.
A launch
contract becomes an industrial partnership
Firefly
says the agreements are worth more than US$1.3 billion. It will
supply 27 sea-level Miranda engines and nine vacuum-optimised Vira
engines for the first three Blacklift vehicles, while providing
engineering support over three years.
AST will also receive a perpetual licence to manufacture both engine
types at a planned US facility. Firefly retains ownership of the
intellectual property and describes the arrangement as non-competitive:
AST intends to launch its own spacecraft, while Firefly remains free to
serve government and commercial customers.
The financial structure reveals how much of the value lies in
knowledge rather than hardware. AST says the licence costs approximately
US$800 million upfront, comprising US$400 million in cash and 10 million
to 14 million Class A shares. Engine deliveries for the first three
rockets cost US$250 million, with a further US$250 million allocated to
engineering support.
Firefly is selling a manufacturing pathway, while AST is buying time,
talent and partial independence from future engine supply.
Blacklift
is more ambitious than the engine purchase suggests
AST’s Blacklift
programme page describes a three-stage vehicle with two reusable
lower stages and capacity for up to five stacked BlueBird satellites per
mission. The company calls Blacklift a “20-ton rocket”, although its
public material does not clearly define whether 20 tonnes is payload
performance, a vehicle-class label or another measure. Until AST
publishes a reference orbit and mass-to-orbit figure, comparisons with
Falcon 9 or other launchers remain provisional.
The engine order implies nine Miranda and three Vira engines per
vehicle, consistent with independent reporting that nine Mirandas will
power the first stage. Firefly rates Miranda at 230,000 pounds-force,
about 1,020 kilonewtons, and says the engine has completed more than 175
hot fires totalling over 4,200 seconds.
Those tests are evidence of propulsion maturity, but neither Miranda
nor Vira has flown. Miranda is also intended for Firefly and Northrop
Grumman’s Eclipse vehicle, whose first launch is scheduled no
earlier than 2027.
AST’s late-2028 target therefore depends on more than receiving
engines. It must design and qualify tanks, structures, avionics,
guidance, stage separation, thermal protection and ground systems. Reuse
adds recovery hardware, mass penalties and another test campaign. A
launch site and recovery architecture have not yet been disclosed.
Buying an engine can shorten development. It cannot buy an
established systems-engineering organisation, flight history or
regulatory licence.
The
economic case is a hedge against launch scarcity
AST’s immediate constellation still depends on external providers.
The company says existing agreements cover more than 100 BlueBird
satellites for cumulative deployment during 2027 and 2028. Blacklift is
therefore not the remedy for near-term service coverage; it is a bet on
replenishment, expansion and additional missions after the initial
network is deployed.
That distinction strengthens the strategic logic. A purpose-built
launcher could reduce payload adapters and integration work, optimise
the fairing around BlueBird’s unusual geometry and allow AST to control
launch dates. If the rocket achieves repeatable reuse, it may lower the
marginal cost of maintaining a large constellation.
It could also support markets AST lists beyond direct-to-device
broadband, including government communications, radar, sensing and edge
computing. These are ambitions, not contracted Blacklift revenue.
The wider competitive implication is significant. Large constellation
operators are increasingly exposed to launch providers that have their
own networks, government priorities or capacity constraints. AST’s move
suggests that control of access to orbit may become part of the
competitive moat for satellite communications, not simply a service
purchased on the open market.
However, dedicated launch capacity only creates value if utilisation
is high. A rocket serving mainly one satellite family has fewer
opportunities to fill schedule gaps with outside payloads, while fixed
costs continue through development delays or deployment pauses.
Firefly gains
capital, scale and a new dependency
For Firefly, the transaction creates a large customer-funded
expansion programme. The company says proceeds will support additional
test bays, long-lead materials and hiring. Higher Miranda and Vira
volumes could lower unit costs and improve the propulsion supply chain
for Eclipse.
The same linkage creates concentration risk. Blacklift and Eclipse
will draw on related engines, expertise and test infrastructure. A
production slip could expose both programmes.
The manufacturing licence eventually reduces AST’s direct dependence
on Firefly, but it introduces a different challenge: reproducing a
high-performance engine with consistent materials, processes and
acceptance testing. Technology transfer is not the same as qualified
serial production.
The
balance sheet can fund the entry price, not guarantee the outcome
AST’s latest filed balance sheet showed US$2.29 billion of cash and
cash equivalents at 30 June 2026. The US$400 million cash element of the
licence alone equals roughly 17% of that amount. The share component
represents approximately 3% to 5% of the Class A shares outstanding in
early August, before considering price movements or other issuance.
Blacklift will compete for funds with satellite production, spectrum,
gateways and external launches. The disclosed US$1.3 billion goes to
Firefly; it does not represent the total cost of facilities, stages,
software, recovery systems and operations.
This is the central investor constraint. Vertical integration can
capture supplier margin and scheduling control, but only after large
upfront expenditure and years of execution risk. Until Blacklift flies
repeatedly, its economics are a forecast rather than an operating
advantage.
Milestones
that will turn the concept into capacity
The first near-term tests arrive quickly. Firefly is due to discuss
the agreement with investors on 10 November, while AST plans a Blacklift
unveiling on 12 November. Those events should clarify vehicle
dimensions, stage architecture, recovery method, launch site, programme
leadership and development budget.
After that come design reviews, Vira hot-fire evidence, acceptance of
the first flight engines, full-stage static fires, recovery
demonstrations and an FAA launch licence. Eclipse’s debut will provide
the first flight evidence for the shared propulsion family.
Investors should also track whether AST’s external launch schedule
remains intact. If more than 100 BlueBirds are deployed as contracted,
Blacklift can mature as a long-term cost and control strategy. If
outside launches slip again, pressure to accelerate an immature rocket
could increase programme risk.
Blacklift is a rational response to a genuine bottleneck, but it is
not yet launch capacity. It is a US$1.3 billion commitment to create
that capacity—and a test of whether constellation economics can support
a second, vertically integrated industrial system.