Starlink satellites seen in the night sky.
Credit: NurPhoto/Getty Images
This week, Amazon announced a new addition to Amazon Leo, the satellite internet initiative formerly known as Project Kuiper. Designed to rival Starlink, Leo Ultra purports to beat the Musk-founded satellite internet service in terms of speeds, with 1Gbps download and 400Mbps upload speeds.
This would make the fastest version of Leo “enterprise-grade,” and a good deal faster than the fastest Starlink subscription option. For reference, Starlink’s “Priority” subscription model offers download speeds of 135 to 310Mbps and upload speeds of 20 to 44Mbps.
That said, Starlink aims to be affordable for a wide range of customers, while the Leo Ultra’s pricing has yet to be announced. The high penetration of Starlink so far could mean that Bezos and company intend to focus on businesses and high-income consumers.
It’s still early days, with the network undergoing testing, but service seems likely to begin operation in the first half of 2026.
Amazon Leo is undoubtedly starting as a latecomer. For reference, Amazon currently has about 150 Leo satellites in orbit, as compared to over 8,000 satellites for Starlink; some estimates say that Starlink operates about two-thirds of all currently active satellites. The FCC has already approved Amazon to launch around 3,500 satellites, and with Starlink, it’s shown a willingness to approve far larger numbers over time.
If Amazon Leo expands to the same extent, the skies could be completely dominated by satellite internet companies delivering a service that already exists. The original pitch for Starlink, versus conventional means of carrying an internet signal, was that it could bring the internet to a wider proportion of the world population; in reality, pricing is such that the service is most logical for globe-trotting Westerners and large work teams in remote areas, like mining and resources companies.
Notably, the first test users for Project Leo are enterprise users.
The base unit for Project Leo.
Credit: Amazon
Among the reasons not to launch these thousands of satellites is the impact of crowding the skies even further than current projects. This poses problems not only by increasing the risk of a catastrophic orbital collapse, as described by the oft-cited Kessler Syndrome, but also by impeding astronomy.
There are some extremely remote communities, particularly island communities, that can realistically only get connectivity with satellite internet, but they are tiny in number and population. With little in the way of upper limits on the number of satellites that Starlink, Amazon, and other enterprises can launch, it’s difficult to see the sky polluted with light, launch-generated carbon, and potential space junk fodder in the name of those communities.
