Amazon's Internet Service Leo Just Crossed 394 Satellites
For five years, the most credible threat to Starlink's dominance over satellite internet has existed primarily on paper, in FCC filings, and in press releases about rockets. As of this week, it exists in orbit. Nearly 400 of them.
Early Thursday morning, a United Launch Alliance Atlas V rocket lifted off from Cape Canaveral carrying 29 more Amazon Leo satellites — the company's 14th launch mission and, notably, the final flight of the Atlas V rocket before ULA transitions its Amazon Leo manifest to the newer Vulcan Centaur. By the time those satellites raised themselves to their assigned altitude over the coming weeks, Amazon Leo's constellation count crossed 394 in orbit — a number that carries more significance than just a milestone.
It's the number Amazon needed to start the internet.
Chris Weber, Amazon Leo's vice president of business and product, posted the confirmation publicly within hours of the launch: enough missions have now been completed for initial service to begin this year. Every future launch, he said, adds coverage and capacity rather than being a prerequisite for turning the network on. That's not a projection or a target date. That's the head of the program saying the infrastructure threshold has been cleared.
Starlink has had this market to itself long enough. That's changing now.
What Amazon Leo Actually Is — And Why the Name Change Matters
Most people who've heard of this project know it by its original codename: Project Kuiper, named after the Kuiper belt and the astronomer Gerald Kuiper. Amazon ran with that name for six years before officially retiring it in November 2025 in favor of Amazon Leo — a name derived from low-Earth orbit, the altitude at which the entire constellation operates.
The rebrand wasn't cosmetic. It came alongside the announcement of three customer terminal product names, a public beta waitlist, and an enterprise preview that put production hardware in the hands of select business customers for real-world testing. Amazon was signaling that Project Kuiper — the engineering codename — was done, and Amazon Leo — the commercial product — was beginning. The name change marked the transition from infrastructure build-out to imminent service launch.
The concept behind Amazon Leo is the same logic that made Starlink successful: instead of a handful of traditional geostationary satellites parked 22,000 miles above Earth, you fill low-Earth orbit — just 390 miles up — with a dense constellation of smaller satellites. The shorter distance to Earth dramatically reduces the latency that made old-school satellite internet frustrating for anything requiring real-time response — video calls, gaming, web browsing — while the density of the constellation ensures continuous coverage as individual satellites pass overhead.
The FCC authorized Amazon to deploy 3,236 satellites in July 2020. In January 2026, the FCC approved an additional 4,500 satellites on top of that, bringing Amazon Leo's total authorized constellation to 7,727 — a number that, when fully deployed, would make it one of the largest satellite networks ever constructed. Amazon is required to have half of the original 3,236 in operation by July 30, 2026, a deadline the current launch pace has been designed around.
How Fast It Is — And What You Pay for It
Amazon has designed three different customer terminals at three different price and performance points, and the spread is notable because it suggests a deliberate strategy of competing across multiple market segments simultaneously rather than targeting a single customer profile the way Starlink's original residential dish did.
The Leo Nano is a compact 7-by-7-inch terminal delivering download speeds up to 100 Mbps — the entry-level option aimed at individual residential customers, smaller businesses, and use cases where portability or small form factor matters more than maximum throughput. The Leo Pro steps up to an 11-by-11-inch antenna and speeds up to 400 Mbps, positioned for homes with heavy usage, small-to-medium businesses, and customers who want headroom beyond basic broadband. The Leo Ultra is the enterprise-grade option, delivering download speeds up to one gigabit per second and upload speeds up to 400 Mbps — the terminal for commercial vessels, remote industrial sites, government customers, and any application where satellite connectivity is mission-critical rather than a fallback.
Amazon hasn't announced official retail pricing for any of these terminals or the monthly service plans attached to them. What the company has consistently said publicly is that affordability is central to the mission — Leo's stated purpose is to bring connectivity to communities and customers beyond the reach of existing networks, and that goal is incompatible with pricing that excludes the people most likely to need it.
Industry analysts tracking the satellite internet market have consistently placed Amazon Leo's expected pricing in direct competition with Starlink's residential tier, which currently runs between $50 and $165 per month depending on plan. Amazon's manufacturing scale and logistics infrastructure — the same supply chain that builds and ships millions of Echo devices and Kindle readers — gives it cost advantages in terminal production that pure-play satellite competitors don't have. Whether those advantages translate into lower hardware prices at launch is something only the official pricing announcement will confirm.
Where Service Is Starting First — And Why the Poles Come Before the Equator
Weber didn't specify which region Amazon Leo will begin service in first, and Amazon hasn't published a country-by-country availability map. But the physics of how the constellation is being built makes the answer fairly predictable: initial coverage begins near the north and south poles and spreads progressively toward the equator as more satellites are added.
This is a function of orbital mechanics rather than a business decision. Satellites in low-Earth orbit don't hover over one location — they move continuously around the planet, passing over different areas with each orbit. Polar orbits provide the most consistent coverage early in a constellation build-out because each satellite sweeps over polar regions on every pass. Equatorial coverage requires more satellites distributed across more orbital planes before the gaps close sufficiently to provide reliable continuous service.
The practical implication for most potential customers: regions at higher latitudes — Alaska, Canada, northern Europe, Scandinavia, Russia, the southern tip of South America, Australia, and New Zealand — are likely to see service availability earlier than equatorial regions including most of South and Southeast Asia, equatorial Africa, and Central America. Amazon hasn't confirmed this explicitly, but the orbital geometry makes it the expected progression.
What Amazon has confirmed is the enterprise preview that started in November 2025, where select business customers have been testing the network with production hardware. That preview has been the real-world stress test between the lab and public launch — the data it's generating on antenna performance, handoff between satellites, latency under real usage conditions, and service stability is directly informing the commercial service configuration.
The 14th Launch — And What the Atlas V Retirement Means
Thursday's mission was the 14th in Amazon Leo's launch sequence and carried specific historical weight beyond the satellite count milestone: it was the last Atlas V rocket to fly an Amazon Leo mission. ULA's Atlas V has been the workhorse of the early constellation build — reliable, well-understood, with a flight record that made it a natural fit for the early missions where launch reliability mattered more than cost per kilogram.
Going forward, Amazon Leo's ULA launches will transition to the Vulcan Centaur — ULA's next-generation rocket that offers meaningfully larger payload capacity and a lower cost-per-launch than the Atlas V. The transition matters for deployment pace: Vulcan can carry more satellites per mission, which directly accelerates the timeline from the current 394 to the 578 that Amazon's FCC license requires for full initial service coverage, and eventually toward the 3,236 that represents the complete authorized constellation.
Beyond ULA, Amazon has secured launch contracts with Arianespace for Ariane 64 flights — which already flew the largest single-mission payload of 32 satellites in February 2026 — and with Blue Origin for New Glenn launches. Jeff Bezos founded both Amazon and Blue Origin, making Blue Origin's role in the constellation build-out an unusual case of a founder's two companies serving as customer and launch provider simultaneously. Amazon has secured over 100 total launches across these providers, representing a multi-billion-dollar commitment to getting the constellation to scale on a specific timeline rather than a best-effort deployment.
Who's Already Signed Up — And What JetBlue Tells You About the Strategy
JetBlue announced earlier this year that it will become the first airline to deploy Amazon Leo for in-flight Wi-Fi starting in 2027. That single partnership reveals more about Amazon Leo's commercial strategy than a dozen press releases about residential broadband.
Aviation is one of the most demanding connectivity environments that exists — aircraft traveling at 500 miles per hour require seamless handoffs between satellites, low and consistent latency for passenger applications, and uptime reliability that airline customers measure against a genuine premium experience standard rather than the "better than nothing" bar that rural residential broadband sometimes operates against. JetBlue choosing Amazon Leo over Starlink Aviation — which already has a significant airline customer base — is a meaningful competitive signal. It suggests that in head-to-head enterprise evaluation, Amazon Leo's technical specifications are already considered a credible alternative to the incumbent.
Maritime customers, remote industrial operations, and government contracts are the other customer categories Amazon has been targeting through the enterprise preview. These segments share a characteristic that residential broadband customers don't: they're currently paying significant money for connectivity that is either very expensive, unreliable, or both — and they have both the budget and the motivation to switch to something better the moment it's available.
What This Actually Means for Starlink
Starlink has approximately 7,000 satellites in orbit, millions of subscribers globally, and four years of head start on commercial operations. That lead is real and it matters — the network effects of an established service with paying customers, refined ground infrastructure, and iterative hardware improvements across multiple terminal generations are not things a new entrant closes quickly regardless of how good the product is.
What Amazon Leo changes is the competitive dynamic that has allowed Starlink to operate largely without a credible peer competitor at the technical level. OneWeb serves enterprise customers but operates a smaller constellation with a different market focus. Telesat Lightspeed has faced funding challenges. Hughes and Viasat operate at much higher orbital altitudes with the latency disadvantages that entails. None of them represent the kind of scale challenge that a company with Amazon's manufacturing capability, logistics infrastructure, AWS cloud backbone, and $10 billion committed to the project presents.
The most direct competitive pressure will likely arrive in two areas. Pricing — Amazon's manufacturing and distribution advantages could allow it to offer competitive terminal prices that pressure Starlink's hardware margins. And enterprise and government contracts — where Amazon's existing relationships through AWS, Amazon Business, and government cloud services give it sales channels that a pure-play satellite company simply doesn't have.
For the average person sitting in a rural area with unreliable broadband, the most important outcome of this competition isn't which company wins. It's that competition between two serious, well-funded providers almost always produces better pricing, better service quality, and faster coverage expansion than either company would deliver alone.
How to Get on the List
Amazon Leo has a public beta waitlist available at leo.amazon.com — the company opened signups alongside the November 2025 rebrand. There are no confirmed details yet on how the waitlist will be prioritized, what the rollout sequence across regions will look like beyond the pole-to-equator orbital logic, or when residential customers in specific countries will receive invitations.
What Weber's statement this week confirms is that the "initial service this year" timeline is real and not contingent on additional launches — the infrastructure is sufficient to begin. Whether "this year" means Q3 or Q4 of 2026, and whether the first commercial customers are residential or enterprise, are the remaining unknowns between now and the moment Amazon Leo's first paying subscribers connect to the internet from a satellite that went up aboard one of those 14 rockets.
Five years of launches, filings, and funding commitments have been building to this. The last Atlas V has flown. The 394th satellite is raising itself to altitude. The waitlist is open.
Amazon Leo is about to turn on the internet.
