Amazon has spent years and committed at least $82 billion in rocket launch contracts to answer a simple question: does Starlink have to be the only option? This year, the answer starts to arrive.

The company's 14th satellite launch, which sent 29 more satellites into orbit aboard an Atlas V rocket from Florida on Thursday, pushed its Kuiper constellation past 390 satellites in orbit. That's enough, according to Amazon's Kuiper chief Chris Weber, to begin initial internet service before the end of 2026.

"We've completed enough launches for initial service this yr," Weber wrote on X after the launch. "Future missions just add coverage and capacity."

What this actually means for internet access

Right now, Starlink is the dominant satellite internet provider, with roughly 10,000 satellites in orbit and service already available to consumers, governments, airlines, and remote operations across the globe. Amazon is still building. But the gap is closing, and competition in this market matters for anyone who lives somewhere that cable and fiber never reached.

Rural America, remote worksites, ships at sea, and aircraft are all places where satellite internet is not a luxury but the only option. A single dominant provider can price accordingly. A second serious competitor changes that math.

Amazon plans to offer Kuiper terminals in sizes ranging from roughly laptop-sized for home consumers to larger, more powerful versions for enterprise and government customers. Initial service is expected to start near Earth's poles, where the satellite coverage is densest, and gradually expand toward the equator as more satellites reach orbit.

The hardware problem no one can ignore

The path forward is not smooth. Amazon has 100-plus rocket launches booked to complete the full 3,200-satellite constellation, but two of its three main rocket options are currently grounded.

Blue Origin's New Glenn rocket exploded on its launchpad last month, destroying the launch tower along with it. Engineers are focused on the engine section to identify the cause. Blue Origin CEO Dave Limp expects launches to resume by year-end, but that timeline depends on what the investigation finds.

The situation is potentially worse for ULA's Vulcan rocket, which is booked to fly at least 40 Kuiper missions. Vulcan encountered a solid rocket motor issue in February and is also sitting idle. It uses the same Blue Origin-built engines as New Glenn, so if the New Glenn explosion is traced back to those engines, Vulcan's return could slip further. ULA says Blue Origin is being transparent as the investigation continues.

That leaves Atlas V, the older workhorse rocket that has now become the primary vehicle keeping Kuiper on schedule. Atlas V is a reliable rocket but has a limited number of remaining flights planned before ULA retires it.

Notably, Amazon has also booked launches on SpaceX's Falcon 9, the same rocket that has built Starlink into a dominant network. Paying a competitor to build a competing system is an uncomfortable arrangement for both companies, but it reflects just how constrained the launch market remains.

The deeper story here is about infrastructure and leverage. Starlink has already demonstrated that satellite internet is not science fiction. It proved useful during the war in Ukraine, serves commercial aviation, and has become a genuine lifeline in disaster response. Amazon entering this market with real scale means the technology gets a second set of investors, engineers, and incentives pushing it forward.

Whether Kuiper can catch Starlink is a long game measured in years and launches. But for the first time, that game has actually started.