Amazon Leo builds its 1000th satellite and promises service launch within weeks

Amazon says it has built the 1000th satellite for its Leo broadband constellation, formerly known as Project Kuiper, and that commercial service is a matter of weeks away. The factory figure looks impressive: the company used to assemble 27 satellites in six months and now builds 27 a week. But only 396 satellites are in orbit. That gap is what cost Amazon its spectrum priority in July.

The factory has outrun the rockets

By simple arithmetic, about 600 built satellites have not yet reached orbit. They sit in warehouses, wait for integration, or wait their turn for a launch. The production problem that held the project back in 2024–2025 appears to be solved. A logistics problem has taken its place: how many rockets Amazon can buy, get onto the launch pad, and fly without failures.

The company uses a mixed pool of launch vehicles, with contracts across several providers, including ULA, Arianespace, Blue Origin, and SpaceX. This diversification protects against dependence on a single contractor, but it makes the schedule fragile. A delay to any new rocket immediately shows up in the number of satellites in space.

What happened with the spectrum

The FCC license required Amazon to deploy half of its planned constellation by mid-2026. The cap was 3236 satellites, so about 1618 by the deadline. The company missed it and in July lost priority on part of its spectrum. According to TNW, the regulator is ready to return it early if Amazon proves it is genuinely expanding the constellation. The exact conditions are cut off in the available portion of the source article, so check the details against the original.

This is an important lever for the business. Spectrum is a limited resource. If a competitor claims the frequencies first, Amazon will have to negotiate coordination in the same bands and share the interference.

Key figures

• 1000 — satellites built
• 396 — in orbit
• 27 per week — current production rate
• 27 per six months — previous rate
• 3236 — planned constellation size under the FCC license
• Service launch — in a few weeks, according to Amazon

Against Starlink, and for AWS

Starlink already has more than 8000 satellites in orbit and millions of subscribers. Amazon will not be able to match its coverage in the coming years. But Amazon has something SpaceX does not: AWS. Satellite connectivity tied to the cloud appeals to enterprise customers, airlines, shipping fleets, telecom operators, and government buyers. They want more than just internet from the sky. They want a link that plugs directly into cloud infrastructure with clear SLAs.

The article's URL mentions Europe. That hints at a regional launch, but the available text gives no details. European customers looking for an alternative to American and Chinese connectivity would see a third major player with its own cloud change their negotiating position.

What to watch next

First: the launch cadence over the next two to three months. If hundreds of satellites reach orbit in that window, the claims about a weekly schedule will become credible. Second: the FCC's decision on spectrum. Restored priority would signal that the regulator accepts the catch-up plan. Third: the terms of the first contracts. Terminal prices and tariffs will show how the company intends to compete with Starlink.

The practical takeaway for operators and integrators is simple. It makes sense to start Amazon Leo pilots once real SLAs and a coverage map exist. For now, these are statements of readiness. For anyone choosing a backup connectivity channel, it is worth planning for both services. That is not because Leo is already better, but because relying on a single satellite provider is costly when it fails.