The Stratolaunch Roc, the Widest Wing Ever Built, and Air-Launching the Talon Hypersonic Off the Back of a Twin-Fuselage Giant

The Stratolaunch Roc is the widest airplane ever built, with a 385-foot wingspan - longer than a football field and wider than a Boeing 747. Originally designed by Microsoft co-founder Paul Allen to air-launch satellites into orbit, it now serves a very different mission: carrying and dropping the Talon A hypersonic test vehicle at altitude so it can be flown, recovered, and flown again. It is, in effect, a reusable “runway in the sky” for testing the hardest speed problem in aerospace.

What Is the Stratolaunch Roc?

The Roc is a twin-fuselage aircraft with six engines and 28 wheels. Its 385-foot wingspan makes it the widest aircraft ever to fly - not the heaviest or longest, but the widest. Parked next to a 747, the 747 looks like a regional jet.

The airframe was built by Scaled Composites in Mojave, California - the same engineering lineage that produced SpaceShipOne. To save cost, the design borrowed heavily from existing hardware: the engines, landing gear, cockpit components, and systems were cannibalized from two used Boeing 747s and hung on an all-new wing.

Only the right-hand fuselage is crewed. The left fuselage is unmanned and exists purely for structure and symmetry. The enormous wing has to be light enough to fly yet strong enough to carry a heavy payload slung beneath its center, between the two bodies.

How Does Air-Launching a Rocket Actually Work?

Air launch is a simple idea: instead of igniting a rocket on the ground and fighting straight up through the thickest part of the atmosphere, you carry the vehicle to altitude beneath an aircraft and release it. The vehicle lights its own engine already around 40,000 feet up, already moving at a few hundred knots, and already above most of the air that would slow it down.

Any pilot who has thought about density altitude understands the principle instinctively. Thin air means less drag and less mass to push through. An air-launched vehicle simply skips the worst of the atmosphere.

This is not a new concept. In the 1950s and ’60s, the North American X-15 - the fastest crewed airplane ever built - was carried aloft under a B-52 and dropped rather than taking off from a runway. SpaceShipOne won the Ansari X Prize in 2004 after being carried up by a mothership called White Knight and released to rocket toward the edge of space. And for decades, Orbital’s Pegasus rocket has been dropped from a modified airliner to place small satellites in orbit. The Roc is the biggest, most ambitious version of a very old idea.

Why Was the Roc Built - and Why Did Its Original Mission Disappear?

The Roc was the vision of Paul Allen, who co-founded Microsoft. Around 2011, Allen funded a company to build the largest airplane on Earth specifically to air-launch rockets into orbit. That effort ran through his Vulcan Aerospace venture, with the airframe built by Scaled Composites.

The aircraft first flew in April 2019 out of Mojave. But Paul Allen died in 2018 - before he ever saw it fly. The original plan lost its champion and its funding direction at almost the exact moment the airplane became real.

The market moved on, too. By the time the Roc flew, small ground-launched rockets had become cheap and frequent, and the satellite-launch business case that justified building the widest airplane in history had quietly evaporated. For a while, the Roc was a 385-foot answer to a question nobody was asking anymore - a recurring reality in aerospace, where the engineering works and the airplane flies while the market shifts underneath it.

How Did Hypersonics Give the Roc a Second Life?

Hypersonic means faster than about five times the speed of sound - Mach 5 and up. At those speeds the physics changes character. Air ahead of the vehicle doesn’t just push back; it heats violently from compression, shock waves stack up, and the vehicle effectively flies inside a blowtorch of its own making. Materials that are fine at Mach 2 can come apart at Mach 6.

The problem the U.S. Department of Defense faced is that traditional hypersonic testing meant bolting a vehicle to a rocket, launching from a range, and destroying everything on every flight. One test, one vehicle, gone. That approach is slow and brutally expensive, and it makes rapid iteration impossible.

Under new ownership, the company operating the Roc - now Stratolaunch - pivoted the entire program. Instead of launching satellites, the Roc would carry a small, reusable hypersonic test vehicle, drop it at altitude, and let it fly a hypersonic profile. That vehicle is the Talon A.

Why Reusability Is the Whole Point

Talon A is designed to fly its hypersonic test, then glide down and be recovered, refurbished, and flown again. The Roc itself lands on a runway and can fly again the following week. Nothing is thrown away - no range countdown, no launch pad, no expendable booster. You take off like an airplane, drop the test article, and land like an airplane.

The timeline, told honestly:

The first powered vehicle, TA-1, made its first powered flight in 2024, following captive-carry and separation tests using a non-powered version to prove the drop itself was safe. TA-1 reached very high speed, gathered data, and came down in the ocean by design. It was not recovered.
TA-2, the reusable vehicle, achieved the milestone that matters. In 2025, Stratolaunch and the Department of Defense reported that Talon A had flown hypersonic, been recovered, and then flown again - a second flight of the same reusable vehicle.

The headline isn’t a top-speed number. It’s that the same machine went hypersonic twice. Reuse is what turns a stunt into a program.

Why This Matters for Pilots and Aviation

Think of the difference between renting an airplane wet for a single flight and owning one you can fly whenever the weather’s good. When every flight is a rental you’ll never repeat, you’re cautious and you learn a little. When you own the machine and can fly it Tuesday and again Thursday, you get good fast, because you’re flying the same aircraft through the same problems until you understand it.

Ground-launched, expendable hypersonic testing is the one-time rental. The Roc and Talon A are trying to make hypersonic testing feel like ownership: fly it, land it, look it over, fly it again. That shift - from expendable to reusable, from launch event to flight operation - is a bet drawn straight from the aviation world and applied to a problem the space world has always solved by blowing things up.

Air launch also gives the program mobility. It isn’t tied to a fixed range or azimuth; a large aircraft can fly to a launch box over open ocean, pick a direction and a time, and release. Combined with a reusable vehicle, that delivers cadence - the ability to fly again and again, which is exactly what hypersonics research has been starving for.

What Are the Limits and Risks?

There is honest cold water here. This is a test bed - not a weapon, not a spaceship, and not a machine that will carry passengers to Tokyo in an hour. Its job is to be a flying laboratory for gathering data on materials, guidance, sensors, and control at hypersonic speed, cheaply and often, so other programs can use what it learns.

The current funding exists because defense hypersonics is a national priority - the Pentagon wants to test hypersonic hardware faster than its adversaries can. That reality cuts both ways:

There is exactly one Roc in the world. When it’s down for maintenance, the whole program waits. A fleet of one has no backup, and the airframe - six engines, two fuselages, an enormous wing that must be inspected foot by foot - is complex to keep flying.
The mission could shift again. The program depends on continued government funding, and this airplane has already been orphaned once when the satellite market collapsed. That risk never fully disappears.
The engineering solves a narrow problem. The leap from a small reusable test dart to anything carrying meaningful payload - or people - is enormous. The Roc is not the front end of hypersonic airliners. What it is, right now, is the best flying test stand for hypersonics anyone has built. That is genuinely important, and it doesn’t need to be more than that to matter.

Who Is Building It?

The airframe came from Scaled Composites, and the original vision came from Paul Allen’s Vulcan Aerospace effort. Today the program is run by Stratolaunch under new ownership, flying out of Mojave and California and Pacific test ranges, with much of the current work done in partnership with the Department of Defense’s hypersonic testing programs. Reporting on these flights comes from Stratolaunch’s own announcements, the Department of Defense, and aerospace trade coverage in outlets such as Aviation Week and Ars Technica.

The widest airplane ever built spent its first years looking like a beautiful mistake - a giant with no job. Then the world’s needs changed, and a runway in the sky turned out to be exactly what the hardest speed problem in aerospace was missing. Sometimes the airplane arrives before the mission it was meant for.

Key Takeaways

The Stratolaunch Roc has a 385-foot wingspan, two fuselages, six engines, and 28 wheels, making it the widest airplane ever to fly.
It was funded by Paul Allen around 2011 to air-launch satellites, first flew in April 2019, but lost its original mission after Allen’s death in 2018 and a shift in the satellite market.
The Roc found new purpose carrying the Talon A hypersonic test vehicle - hypersonic meaning Mach 5 and above.
Reusability is the breakthrough: in 2025, the reusable TA-2 flew hypersonic, was recovered, and flew again, following TA-1’s first powered flight in 2024.
The system is a reusable flying test bed for defense-funded hypersonics research - not a weapon, spaceship, or passenger aircraft - and its biggest vulnerabilities are being a fleet of one and dependence on continued government funding.