AST SpaceMobile Taps Firefly Engines for 20 Ton Blacklift Rocket in $1.3 Billion Deal

AST SpaceMobile has spent years arguing that ordinary smartphones should connect directly to satellites, with no dish, terminal or special handset. The Texas company designs and operates BlueBird satellites in low Earth orbit and sells capacity to mobile network operators. Its proposition depends on scale: more satellites mean more coverage, more bandwidth and more paying partners. That makes access to launch a strategic question rather than a procurement detail.

So far AST has relied on other people's rockets. BlueBirds have flown on SpaceX's Falcon 9, Blue Origin's New Glenn and India's LVM-3. Space.com reported on November 6 that AST has launched 13 BlueBirds to date, including eight Block 2 variants whose antennas cover 2,400 square feet, or 223 square metres, which the company calls the largest ever deployed in low Earth orbit by a commercial satellite.

Relying on the market has costs. Reuters reported on November 6 that roughly a dozen BlueBird satellites have launched since 2024, and that in April a New Glenn issue left one BlueBird in a lower-than-planned orbit, from which it re-entered days later. Satellite operators have struggled to book rides as SpaceX, and in future Rocket Lab, dedicate manifests to their own constellations. Launch is becoming a seller's market.

Against that backdrop AST announced on November 6, 2026 that it will build its own rocket. The vehicle is called Blacklift, it is designed for reusability, and AST classifies it as a 20-ton rocket, meaning about 20 tons to low Earth orbit. The company targets late 2028 for a debut flight and will buy the engines from Firefly Aerospace under an agreement valued at more than $1.3 billion.

Key Facts

Reuters reported on November 6 that Firefly will build and supply 27 Miranda engines and nine Vira engines for the first three Blacklift rockets, and that the agreement grants AST a licence to eventually manufacture the engines itself. Spaceflight Now reported on November 7 that the split implies nine Miranda engines on each first stage and three Vira engines on each upper stage. Firefly's medium-lift Eclipse rocket uses seven Miranda and one Vira.

The money is structured in three parts. Spaceflight Now reported on November 7 that the $1.3 billion comprises $800 million upfront from AST, half in cash and half in Class A common stock, plus $250 million for the 36 engines and $250 million over three years for engineering services. Firefly Aerospace said in its announcement on November 6 that the stock portion is subject to lock-up periods. The equity component covers roughly 10 million to 14 million AST Class A shares.

The market reaction was lopsided. Reuters reported on November 6 that Firefly shares rose about 18 per cent after the announcement while AST stock was up about 1 per cent. Firefly, based in Cedar Park, Texas, will use the cash to expand hiring, buy long-lead materials and add test bays at Rocket Ranch, its manufacturing and testing hub in Briggs, Texas, about 20 miles from its Austin-area headquarters.

The engines are not yet proven in flight. Firefly Aerospace said on November 6 that Miranda has performed more than 175 hot fire tests totaling more than 4,200 seconds, including a flight-like test at 110 per cent of the 230,000 pound-force engine's performance that ran 20 seconds beyond the nominal 206-second flight duration. Miranda and Vira scale technology from Firefly's Reaver and Lightning engines. Neither has flown to orbit.

Blacklift's published specifications remain thin. Spaceflight Now reported on November 7 that height, propellant tank sizes and launch site are unknown, and that AST has not explained how it will go from rendering to launch in about two years. The same report places Blacklift in the class of SpaceX's Falcon 9 and United Launch Alliance's Vulcan, while Space.com noted Falcon 9 hauls about 25 tons to low Earth orbit against 20 tons for Blacklift.

Analysis

The strategic logic is easy to state. A constellation is only as valuable as the schedule that fills it. AST says it has over 100 launch agreements toward a total deployment of 248 satellites for its direct-to-phone network, according to Reuters on November 6. Those agreements keep the company moving while Blacklift is developed. Owning a rocket, and eventually the engine inside it, turns a dependency into an asset.

What this really means is that AST is buying optionality, not a finished rocket. The licence gives it the right to build Miranda and Vira engines itself after the first three vehicles, a manufacturing capability it does not have today. Firefly keeps full ownership of the engine intellectual property, and both companies call the arrangement non-competitive because AST launches only its own satellites while Firefly serves government and commercial customers. That leaves Firefly free to sell Miranda and Vira to outside buyers, including its own Eclipse vehicle, without turning AST into a competitor.

The timeline is the weak point. Payload Space analysis cited by Spaceflight Now found that it historically takes about six years from development start to a successful first launch, and AST promises roughly two. Vira is still in development and neither engine has reached orbit. Firefly's own public estimate for Eclipse's debut is 2027, so a 2028 Blacklift flight would require qualification, integration, launch site work and approvals to compress into an unusually short window.

The bigger picture here is that launch scarcity has become a vertical integration problem. SpaceX has said it will phase out Falcon 9 and Falcon Heavy once Starship is operational, and has floated 100,000 Starlink V3 satellites and a million Starmind AI data centres, which would absorb enormous capacity. Cowboy Space unveiled its own Big Iron rocket ten days before AST's announcement, also aimed at 2028. Rides are being reserved years ahead.

Why It Matters

For AST the deal is about network economics. CEO Abel Avellan described Blacklift as the springboard to significantly expand the company's total addressable market, providing regular unconstrained access to orbit and, in his words, the most favorable owners' economics in the industry. Vertical integration is meant to lower cost per satellite and remove scheduling risk. Avellan also called Firefly one of a few companies in the United States with a proven liquid engine of this scale.

For Firefly the deal is validation and capacity. CEO Jason Kim said the agreement supercharges Firefly's production system, with more test bays at Rocket Ranch, long-lead materials bought ahead and a bigger team of engineers and technicians. Miranda is also the heart of the in-development Eclipse rocket, so the AST order funds expansion of the same engine line and adds volume that should help both programs.

For the wider industry the message is that operators no longer trust the launch market to serve them. AST has 13 BlueBirds in orbit and contracts for more than 100 more across 2027 and 2028, Space.com reported on November 6. If a direct-to-phone network needs 248 satellites, the gap between a two-year and a six-year rocket program is measured in subscribers and revenue.

Next Up

AST holds an unveiling event in New York City on Thursday, November 12, and reports third-quarter earnings on Monday, November 9 at 5 p.m. Eastern. Firefly reports on November 10 at 5 p.m. Eastern. Those appearances should add detail on Blacklift's design, its launch site and the pace of engine production.

The milestones are concrete. Miranda and Vira need to fly, Eclipse needs to reach orbit, and AST needs to show that the first Blacklift stage is more than a rendering. The licence and engineering support cover three years, which sets a rough deadline for AST to stand up its own engine manufacturing. If the schedule slips, the 100-plus launch agreements remain the fallback.