Starlink wins this comparison in 2026, because it is the only one of the two a household can actually order. At OrbitalIntel we track both constellations from regulatory filings and from what each company publishes itself.
Amazon Leo is the permanent consumer brand for the satellite network Amazon started as Project Kuiper, and the Federal Communications Commission (FCC) now uses the Leo name for the constellation in its own orders. The two services fly at similar altitudes and target similar latency. Amazon Leo vs Starlink comes down to two things right now, how many satellites each has in orbit and whether you can buy the service at all.
Amazon Leo vs Starlink at a Glance
Starlink is available to order today across most of its coverage map, and Amazon Leo is not yet open to residential customers anywhere. That single difference decides the comparison for anyone who needs a connection this year, before speed or price enters the argument.
| Criterion | Amazon Leo (Project Kuiper) | Starlink |
|---|---|---|
| Satellites in orbit as of 2026 | More than 392 production satellites across 19 missions | Thousands, the largest fleet ever flown |
| Authorized fleet | 3,232 Gen1 satellites, all due in orbit by July 30, 2029 | Operating and expanding under its own FCC authorizations |
| Who can buy it | Selected enterprise preview customers only, plus a public waitlist | Households, businesses, boats, aircraft, in most covered markets |
| Latency | Low, tens of milliseconds, same low-orbit physics | Low, tens of milliseconds |
| Published consumer price | None. Amazon has published no plan or hardware price | Plan and hardware kit prices published by address |
| Launch vehicles | Atlas V, Falcon 9, Ariane 6, Vulcan Centaur, New Glenn | Falcon 9 and Starship, both owned by SpaceX |
| Verdict | Wait-list it. Do not plan a 2026 connection around it | The only workable choice today for a household off the wired grid |
Our Project Kuiper explainer sets both networks alongside OneWeb in a three-way orbit table. The two-way race turns on the spectrum condition the FCC attached to Amazon Leo in June 2026.
Will Amazon Leo Compete with Starlink?
Amazon Leo will compete with Starlink, and it is the only competitor with the capital to build a fleet of comparable size. Amazon told the FCC in its milestone application that it has invested more than $10 billion in deploying the system, alongside spending on manufacturing capacity and dedicated launch infrastructure. That is the scale of commitment a rival needs to matter, and no other Western operator has matched it.
The competition is real for buyers even before Amazon Leo opens. A second large low-orbit network gives a household in a Starlink-waitlisted area a second chance at service instead of no chance. It also gives regulators an alternative they lack today. In its June 2026 order the FCC noted that only one operator, SpaceX, provides low-orbit broadband to American consumers.
Amazon has one advantage Starlink cannot copy. Amazon already runs a retail and logistics operation reaching tens of millions of households, plus its own cloud infrastructure through Amazon Web Services. That means it can sell and ship a terminal through channels it already owns. Whether that turns into a cheaper monthly bill is unannounced, and worth nothing to a buyer until it is.
Amazon Leo vs Starlink on Number of Satellites
Starlink flies thousands of operational satellites, and Amazon Leo has launched more than 392 production satellites across 19 missions as of 2026. A low-orbit satellite passes overhead in minutes, so a network needs many of them spread across many orbital planes before any single address gets a connection that does not drop. Our LEO satellites explainer covers why that handoff requirement forces the fleet count so high.
The gap is easier to read from Amazon’s own filings than from a headline count. Amazon applied to the FCC for a milestone extension on January 30, 2026, in ICFS file SAT-MOD-20260129-00065. That filing reported 180 satellites launched and projected roughly 700 by July 30, 2026, about 21 percent of the Gen1 authorization. The license requires 1,616 satellites by that July 2026 date and the full 3,232 by July 30, 2029.
Amazon says it has scaled manufacturing to 30 satellites a week to close that gap. At a sustained rate like that the arithmetic works, but production is not the constraint that has been binding, and the next section explains which one is.
What the FCC’s June 2026 Waiver Costs Amazon Leo
The FCC granted Amazon Leo a limited waiver on June 5, 2026 and attached a condition that costs Amazon Leo its spectrum priority for up to 20 months. Any Gen1 satellite not deployed and operational as of July 30, 2026 temporarily loses the earlier-round priority status Amazon won in the 2020 Ka/Ku-band and 2021 V-band processing rounds. That status runs out to March 30, 2028, or until Amazon Leo deploys and operates half its constellation, whichever comes first.
Losing earlier-round status is not a paperwork detail. During that window Amazon Leo has to coordinate with the operators it would otherwise have outranked, or operate on a non-interference basis toward them. A newly launched Amazon Leo satellite gives way instead of being given way to. For a buyer, that is the difference between a network that can defend its capacity and one working around a rival.
Amazon can shorten the penalty to 15 months, ending October 30, 2027. It has to certify to the FCC that it built every satellite needed for the 50 percent milestone and secured enough launches to fly them. SpaceX filed in opposition to the waiver and argued the undeployed portion of the fleet should go back into a new processing round instead. The FCC disagreed and kept Amazon Leo’s July 30, 2029 final deadline intact, so the build-out schedule survives.
How Fast Will Amazon Leo Be?
Amazon has published a speed figure for exactly one Amazon Leo terminal, and it is not the one a household would get. The Leo Ultra is an enterprise antenna. Amazon’s own announcement rates it at up to 1 Gbps download and up to 400 Mbps upload, and calls it the first commercial phased array antenna to support gigabit speeds. Amazon also names two smaller terminals, the Leo Nano and the Leo Pro, without publishing residential plan speeds for either.
An enterprise antenna’s ceiling tells you very little about what a home connection will do. Satellite capacity over any area is shared, so the number a subscriber sees depends on how many other subscribers are pulling from the same satellites at the same moment. Starlink publishes its own plan speed ranges by address on the Starlink residential page, and those are the figures to compare against once Amazon posts an equivalent.
Until Amazon publishes residential speeds at leo.amazon.com, any Amazon Leo speed comparison is a comparison against a ceiling rather than a service. The Leo Ultra’s 1 Gbps shows the hardware works. It is not a plan a household can buy.
Amazon Leo vs Starlink on Latency
Latency is close to a tie, because both networks fly a few hundred kilometres up and both land in the tens of milliseconds. That is the whole reason low orbit was worth the extra satellites. A geostationary satellite sits nearly 36,000 kilometres out and adds roughly 600 milliseconds of round-trip delay, which made older satellite internet unusable for video calls and gaming.
Orbit height sets the floor on delay, and it is the same floor for both. The ground segment separates the two in practice, on three counts:
- How many gateway stations feed traffic into the internet backbone, which caps how much traffic the network can land.
- Whether satellites relay to each other with laser links instead of routing back down, which keeps a connection working over oceans and empty terrain.
- How much spare capacity sits over a busy area, which keeps evening speeds from collapsing.
Our how satellite internet works explainer walks through that path end to end.
Starlink has years of operating data behind those choices and Amazon Leo does not yet. Our Starlink explainer covers the dish, the laser links, and the gateway path in detail. So the fair reading in 2026 is that Amazon Leo should match Starlink’s latency on physics and has not demonstrated it under load.
How Much Will Amazon Kuiper Cost?
Amazon has published no consumer pricing for Amazon Leo, which means every dollar figure circulating for it is an estimate someone made up. That includes the monthly plan, the terminal, and any Prime bundling. Amazon runs a free waitlist at leo.amazon.com with no deposit and no commitment, and that page is where a real price appears first.
Starlink’s cost structure is published and has two parts. A customer pays once for the hardware kit, the dish and router they own outright, then a recurring monthly plan on top. Both numbers vary by country and by plan tier, and Starlink shows the current figures for a specific address on its order page.
The pricing gap matters for a decision you might be making right now. A published price you can budget against beats an unpublished one that might be lower, especially when the unpublished service has no residential availability date attached to it.
Rockets Are the Binding Constraint on Amazon Leo
Amazon told the FCC that launch provider delays and shortages of launch availability caused the backlog behind its missed milestone. That is the clearest structural difference between the two companies. SpaceX builds Starlink satellites and launches them on its own Falcon 9 and Starship vehicles, so it adds capacity at whatever rate it chooses to fund. Amazon buys launches from other people and waits for their schedules.
Amazon spread its manifest deliberately rather than betting on one rocket:
- Atlas V and Vulcan Centaur from United Launch Alliance, which carried the early production missions.
- Falcon 9 from SpaceX, the direct competitor in satellite broadband, bought because schedule mattered more than optics.
- Ariane 6 from Arianespace, which gives Amazon a European pad outside United States range congestion.
- New Glenn from Blue Origin, the vehicle with the most Amazon Leo capacity per flight and the least flight history.
Spreading the manifest protects Amazon against any single rocket standing down. It still leaves the deployment rate outside Amazon’s control. So watch the deployment conditions in the FCC’s June 2026 order rather than the factory.
Who Is the Biggest Competitor to Starlink?
Amazon Leo is Starlink’s biggest competitor by capital committed and by satellites authorized, but not by satellites currently serving customers. Amazon holds authorization for 3,232 Gen1 satellites. The FCC’s Space Bureau separately authorized an Amazon Gen2 system of 3,212 satellites and a polar system of 1,292, with milestone dates in February 2032 and February 2035.
OneWeb, now part of Eutelsat, currently flies more satellites than Amazon Leo does. It aims at enterprise, maritime, aviation, and government buyers rather than households, so it does not compete for the same customer even where it competes for the same spectrum. Our best satellite internet providers roundup ranks all of them against the geostationary options as well.
The other serious contender is state-backed rather than commercial. China’s Guowang and Qianfan constellations are being built at national scale with their own launch capacity, and we cover them in our China satellite constellations explainer.
Who Should Not Wait for Amazon Leo
Three groups should order what exists today rather than wait.
- Anyone who needs a working connection in 2026. Amazon Leo has no published residential availability date, and the FCC record shows a fleet roughly a fifth of the way to its interim milestone. Order Starlink if it serves the address, and join the Amazon Leo waitlist as a backup.
- Anyone with fiber or cable available at the property. A wired connection beats either satellite network on speed, price per gigabit, and consistency. Neither service is worth switching to when a ground line is already there.
- Anyone outside the first rollout markets. Amazon Leo’s early coverage follows the satellites, which pass over higher latitudes most often, so equatorial regions come last. Check the Starlink availability map for the address instead of waiting.
What Would Change Our Answer
Three specific events would flip this verdict, and none of them is a rumour or a leak.
- Amazon publishes residential pricing and opens general availability at leo.amazon.com. A published monthly price with a real order flow makes Amazon Leo comparable rather than hypothetical, and if it undercuts Starlink meaningfully it becomes the better buy for cost-driven households.
- Amazon certifies to the FCC that it has built and procured launches for half the Gen1 fleet. That certification shortens the spectrum priority penalty from 20 months to 15 and signals the launch bottleneck has eased.
- Amazon Leo’s deployed count crosses roughly 1,600 operational satellites. At that point it holds continuous coverage over most populated latitudes, and the coverage argument for Starlink stops being decisive.
If you are weighing Amazon Leo vs Starlink for a specific address, check the Starlink coverage map today and join the Amazon Leo waitlist the same afternoon. The waitlist costs nothing, and the coverage answer decides whether you get connected this year.