SpaceX has taken a big regulatory step toward a much larger Starlink: it filed an application with the U.S. Federal Communications Commission (FCC) to operate a 100,000-member constellation of “Gen3” satellites in low Earth orbit (LEO). The most telling part is the size. According to FCC application details shared by astronomer and satellite tracker Jonathan McDowell, each Gen3 satellite is listed at 4,400 to 5,500 pounds (2,000 to 2,500 kilograms). That is not incremental. It is a different class of satellite, designed to broadcast broadband from a far more dense orbital presence.

This is also a scale-up from where SpaceX already is. SpaceX currently operates nearly 10,800 internet satellites in LEO and has FCC approval for about 4,000 more. The new request would dramatically expand Starlink’s orbital footprint by adding a 100,000-member system to an already substantial base. In other words, this is not a “future project” filing in the abstract. It is an explicit bid to expand the size of the LEO internet layer, and the FCC is the gate SpaceX has to get through.

McDowell’s notes add the practical engineering context executives should care about: the Gen3 satellites are described as covering an area of 3,230 to 4,300 square feet (300 to 400 square meters) with their solar arrays extended. The application also lists orbital parameters, with orbits at 320 to 480 km at various inclinations. Those numbers matter because they translate into how much sky resources are consumed, how often satellites can be visible, and how they populate different orbital planes. And because Gen3 is materially larger than the satellites SpaceX launches today, it changes what launch capacity the company can rely on for deployment.

For perspective, the Starlink satellite version SpaceX is launching these days, known as the V2 Mini, weighs about 1,760 pounds (800 kg) and covers about 1,250 square feet (116 square meters), according to SpaceflightNow. The V2 Mini is launched on SpaceX’s Falcon 9 in fleets of 29 satellites apiece. Gen3, however, is so big that McDowell’s framing suggests it will probably need to fly on Starship, SpaceX’s giant, super-heavy-lift rocket. That implication is strategic. If Starship deployment becomes the bottleneck, then regulatory approval is only step one. The capital allocation and schedule risk move from “can the satellites be approved” to “can the launch architecture actually deliver the constellation at the requested scale.”

There is also the market and regulatory backdrop around what “scale” means in LEO. A number of companies are building large constellations, including Amazon and Blue Origin, and many more have plans in the near future. At the same time, a variety of people and organizations have objected to how crowded Earth orbit is becoming, voicing concerns about the effects on astronomical observations, wildlife, Earth’s atmosphere, and humanity’s enjoyment of the night sky. This is the collision point executives should track: megaconstellations are not being evaluated only on connectivity performance, but also on externalities. When a plan grows from thousands to 100,000 satellites, the “how crowded is too crowded” debate moves from background noise to foreground policy.

SpaceX is also signaling that this is part of a broader ambition beyond broadband. The source notes that the company aims to build a million-strong AI megaconstellation called Starmind. Musk wrote in a February 2026 update about the planned AI constellation: “Launching a constellation of a million satellites that operate as orbital data centers is a first step towards becoming a Kardashev II-level civilization, one that can harness the sun's full power, while supporting AI-driven applications for billions of people today and ensuring humanity's multi-planetary future.” Even if you ignore the visionary framing, the operational point stays grounded: scale in space can become a platform strategy, not just an infrastructure project. Bigger constellations mean more data, more compute-adjacent infrastructure, and more leverage in the ecosystem.

For decision-makers at competitors, regulators, and infrastructure partners, this FCC filing is a clear signal of where the pressure will be applied. SpaceX is already operating thousands in LEO and has approval for more, and now it is pushing a Gen3 expansion that is both heavier and physically larger per satellite, with a dramatically enlarged orbital footprint. If you are assessing your own risk in LEO spectrum access, launch scheduling, satellite manufacturing capacity, or policy engagement, this is the moment to focus: the constellation size is no longer a theoretical number. It is a concrete regulatory request with knock-on effects for launch architecture, orbital congestion, and how much scrutiny the entire sector can expect.