On the sidelines of the Tashkent International Investment Forum this week, a joint venture owned by Uzbekistan’s state telecom operator signed an agreement to resell internet delivered from space. The American company on the other side of the table was Amazon, and the deal puts its satellite service in direct competition with Elon Musk’s Starlink across a region Starlink has been entering first.
UZ-SAT, a joint venture under Uztelecom, will serve as the authorized distributor of Amazon Leo in Uzbekistan, the companies said. The arrangement gives Amazon a local partner that already covers 96 percent of the country’s populated areas and operates 378,000 kilometers of fiber, and it gives UZ-SAT a way to reach the parts of the country that fiber cannot: the mountains and the desert.
Amazon Leo is the constellation formerly known as Project Kuiper, a network of more than 3,000 low-Earth-orbit satellites backed by a ground system of gateway antennas and dedicated fiber. Amazon has said it plans a full commercial launch in 2026, and analysts have put its total investment in the project at between $10 billion and $20 billion. The unit traces its origins to 2019, when Amazon first began building the satellite operation that became Kuiper and, later, Leo.
Trevor Vieweg, the global head of business for Amazon Leo, framed the deal around reach rather than rivalry. “We’re excited to bring Amazon Leo to Uzbekistan, delivering high-performance connectivity to communities and businesses that need it most,” he said, adding that pairing the satellite technology with local expertise would “open new doors for economic growth, education, and innovation.”
The product lineup gives UZ-SAT a ladder of offerings. The Leo Nano terminal is aimed at consumers, with speeds up to 100 megabits per second. The Pro serves heavier bandwidth needs, and the Ultra, an industrial-grade antenna, can deliver up to a gigabit. The tiering is meant to serve everyone from a household in a remote village to an enterprise in the capital.
Uzbekistan is not Amazon’s first stop in the region. Last September the company announced an entry into Kazakhstan, naming Kazakhtelecom as its local partner. That deal established the pattern the new one follows: a state-linked telecom operator handles distribution and support, while Amazon supplies the satellites, the terminals, and the network.
The timing matters because Starlink has been moving through the same neighborhood. SpaceX’s service has already launched in Kazakhstan, Tajikistan, and Kyrgyzstan, according to local reports, giving Musk’s network a head start in Central Asia. Amazon is arriving later and leaning on incumbent operators who already own customer relationships and fiber routes.
For Uzbekistan, the appeal is practical. A country where a large share of the population lives outside the major cities has struggled to extend terrestrial broadband into mountainous and arid terrain, where laying fiber is slow and expensive. A low-Earth-orbit service can reach those areas without the trenching, which is why the government framed the agreement as a way to close the rural connectivity gap.
For Amazon, the appeal is distribution. A satellite constellation is a fixed cost: the network is built whether a million customers or ten million use it. Amazon’s own reasoning has long held that the service makes sense only at scale, and distributor agreements with incumbent operators are the fastest way to reach subscribers in markets where building a retail footprint from scratch would take years.
The satellite network also fits a wider Amazon strategy. The company has spent years building cloud, retail, and logistics businesses that reinforce one another, and connectivity is the substrate they all run on. Selling bandwidth through a national operator gives Amazon a foothold in a market without the cost of building a terrestrial network, while its terminals and customer support ride on UZ-SAT’s existing reach into most of the population.
Uzbekistan’s geography concentrates the problem the partnership is meant to solve. Much of the country’s terrain is desert, and its eastern edge rises into mountain ranges where villages sit hours from the nearest fiber line. Satellite coverage can reach those communities without the months of engineering that trenching requires, which the government has cited in pitching the deal as a step toward universal access.
The rivalry with Starlink is direct but lopsided for now. Starlink has a working constellation, millions of subscribers, and years of launch cadence behind it. Amazon’s constellation is still ramping toward commercial service, and the company must close that gap while Starlink keeps signing new countries. Analysts said the contest in Central Asia will come down to price, terminal cost, and how quickly Amazon can turn distributor deals into actual subscribers.
Neither company is framing the fight as a zero-sum game in public. But the region is a test of a larger question hanging over Amazon’s satellite bet: whether there is room for a second low-Earth-orbit network to build a real business against a competitor that got there first and has not slowed down.
