SpaceX has applied to the U.S. Federal Communications Commission for permission to operate a 100,000-satellite “Gen3” system in low Earth orbit, with individual spacecraft reportedly weighing between 2,000 and 2,500 kilograms and spreading 300 to 400 square meters once their solar arrays are extended.
The proposed architecture is even more specific than the headline number. Jonathan McDowell’s constellation database, compiled from the filing, lists 20 orbital shells containing 5,000 satellites apiece: ten between 323 and 327.5 kilometers and ten between 473 and 477.5 kilometers, at inclinations ranging from 26 to 96.9 degrees.
For comparison, SpaceX currently operates about 10,800 Starlink satellites in orbit. The new application would therefore describe a network more than nine times the size of the fleet already circling Earth, although the filing reportedly calls it Gen3 rather than explicitly naming it Starlink.
The filing is real, but the 1,600-launch figure is not established
The size and mass of the proposed satellites point toward Starship. A Falcon 9 currently carries about 29 V2 Mini Starlink satellites at a time, and those spacecraft weigh roughly 800 kilograms each. A Gen3 satellite weighing as much as 2,500 kilograms would be more than three times heavier.
What the public material does not establish is that Starship would carry precisely 60 Gen3 satellites on every flight. Without a verified payload configuration, deployment mechanism, reserve margin and final Starship performance figure, dividing 100,000 by 60 and presenting the result as more than 1,600 required launches creates a precise number from an unverified assumption.
The defensible calculation is still enormous. At 2,000 to 2,500 kilograms per satellite, the complete constellation would contain between 200 million and 250 million kilograms of spacecraft — 200,000 to 250,000 metric tons — before counting dispensers, structural hardware or anything required to transport them into orbit.
SpaceX has shown that launch cadence can change what once looked industrially impossible. Its reused Falcon 9 boosters now cycle through Starlink and commercial missions at a rate examined in Space Daily’s account of a booster completing its seventeenth flight. Starship, however, has not yet demonstrated a comparable operational freight system.
Twenty shells occupy two very different altitude bands
The lower half of the proposed constellation would orbit within a narrow 4.5-kilometer band, from 323 to 327.5 kilometers. The upper half would repeat the same inclination pattern between 473 and 477.5 kilometers.
That means the entire constellation would not sit below the International Space Station. The ISS normally operates near 400 kilometers: the ten lower Gen3 shells would pass beneath it, while the ten upper shells would orbit roughly 70 to 80 kilometers above it.
Operating near 325 kilometers offers shorter signal paths and causes disabled spacecraft to lose altitude faster than satellites at conventional Starlink heights. The trade-off is atmospheric drag. Even the thin gas at those altitudes continuously removes orbital energy, requiring more frequent reboosts and making satellite lifetime more sensitive to changes in the upper atmosphere.
The engineering problem becomes progressively harder below 300 kilometers, where drag rises sharply. China, Japan and Europe have been testing ways to remain in that region, including the sustained very-low-orbit missions described in Space Daily’s examination of China’s new VLEO alliance.
Gen3 and Starmind are separate proposals
The Gen3 application describes an extraordinarily large communications constellation. Its scale may eventually support uses beyond ordinary consumer broadband, but the filing should not automatically be treated as proof that all 100,000 satellites are orbital data centers or a dedicated AI network.
SpaceX has discussed a separate concept involving as many as one million computing satellites. The company described that plan publicly in February 2026 and Elon Musk confirmed the name “Starmind” in June.
That proposal included language about solar-powered AI computing and a future civilization capable of using a star’s energy. The Gen3 filing arrived later with different satellite numbers, masses and orbital bands.
The projects could ultimately share launch vehicles, communications infrastructure or manufacturing capacity. That remains a possible relationship rather than an established sequence in which Gen3 is built first to provide the downlinks for Starmind.
The astronomy effect depends on altitude, orientation and sunlight
A network containing 100,000 large spacecraft would create a serious new burden for optical astronomy. Satellite trails already cross long-exposure images, and work prepared for the Vera C. Rubin Observatory has shown that particularly bright trails can leave effects extending beyond the obvious streak itself.
Lower altitude does not simply mean that satellites remain illuminated for more of the night. A study of the low-Earth-orbit satellite population found that low-orbiting spacecraft can spend several hours in Earth’s shadow around the middle of the night, depending on season and latitude.
The more difficult period is twilight. Modeling of satellites at 350 kilometers found that their effect would be more severe during astronomical twilight but less severe during full darkness than comparable spacecraft at 550 kilometers. When illuminated, a lower satellite is closer to the observer and moves across the detector more rapidly.
The proposed 300-to-400-square-meter area raises an obvious visibility concern, but area alone does not determine brightness. Surface materials, spacecraft attitude, solar-array angle and the direction in which reflected light is sent can make two similarly sized objects look very different from the ground.
Radio astronomy faces a separate problem. LOFAR observations have detected unintended electromagnetic radiation from Starlink satellites between 110 and 188 MHz, including emissions from frequencies not being deliberately used for communications. Multiplying the spacecraft population increases the number of potential transmitters moving through a radio telescope’s field of view.
Low orbit makes space weather an operational variable
Atmospheric density at several hundred kilometers changes as the Sun heats and disturbs Earth’s upper atmosphere. A satellite can encounter substantially more drag during a geomagnetic storm than mission planners expected from quiet conditions.
SpaceX experienced that directly in February 2022. A Falcon 9 released 49 Starlink satellites at an altitude of roughly 210 kilometers shortly before geomagnetic activity increased atmospheric drag. According to a later analysis of NORAD tracking data, 38 of the 49 satellites were eventually lost.
That incident does not mean a storm would produce the same result at 325 or 475 kilometers. The failed satellites were newly deployed more than 100 kilometers below the proposed lower Gen3 shells and had not yet completed their orbit-raising maneuvers. It does show why a constellation operating close to the atmosphere requires accurate forecasting, propulsion reserves and rapid autonomous responses.
Reentry also becomes a question of scale. A constellation containing 200,000 to 250,000 metric tons of hardware would eventually return material to the atmosphere as satellites failed or reached retirement. The chemical effects of sustained spacecraft reentry remain an active area of research, and the composition of the proposed Gen3 vehicles has not been publicly described in enough detail to calculate the result.
An application is not an authorization
The FCC has previously granted SpaceX less than it requested. In its 2022 Gen2 decision, the agency authorized 7,500 of the nearly 30,000 proposed satellites, deferred the remainder and imposed conditions covering debris, collision reporting, scientific missions, radio astronomy and environmental concerns.
A 100,000-satellite request begins another regulatory process rather than guaranteeing that 100,000 spacecraft will fly. Other operators, astronomy organizations, environmental groups and government agencies can submit technical objections before the FCC decides whether to approve, narrow, defer or reject parts of the system.
Starship remains the physical test behind the paperwork. The rocket must move from developmental flights to repeatable orbital freight operations, and SpaceX must establish how many Gen3 spacecraft can actually be carried, released and replenished on each mission. Until those numbers exist, any total launch count is an estimate built on assumptions.
For now, the constellation exists as rows in a filing: 5,000 satellites at one altitude, another 5,000 half a kilometer higher, repeated until the count reaches 100,000. If those rows ever become hardware, the first visible sign will arrive at twilight, when moving points of reflected sunlight begin crossing the darkening sky in numbers no previous generation has had to plan around.
