Amazon's satellite internet service — Amazon Leo (formerly Project Kuiper) — finally has a launch window. Beta in late 2026, broader US service in 2027. The natural question every rural homeowner is asking: do I wait for Leo or get Starlink now? Here's the honest call after digging through Amazon's announcements, FCC filings, and the satellite-tracker data.
If you need internet at your house in 2026: get Starlink. It's available today, it works, and Leo doesn't have commercial residential service yet.
If you can wait until late 2027: watch Leo. The pricing target ($50–$100/mo vs Starlink's $120) and Prime bundle potential could shift the math significantly.
If you're already on Starlink and considering switching: wait at least 12 months past Leo's beta to see real-world performance, churn rates, and customer support quality. Amazon has run beta-quality products before and the early experience can be rough.
| Amazon Leo | Starlink | |
|---|---|---|
| Tech | LEO satellite (~370 mi orbit) | LEO satellite (~340 mi orbit) |
| Status | Beta late 2026 / early 2027 | Commercial, available now |
| Satellites in orbit | ~331 (May 2026) | ~6,800+ (active) |
| Full constellation | 3,236 planned | ~12,000 planned |
| Speed (residential) | Up to 400 Mbps (Leo Pro) | 50–250 Mbps typical (Standard) |
| Latency | 30–50 ms (target) | 20–40 ms typical |
| Hardware cost | ~$400 target (Leo Pro) | $349–$499 standard kit |
| Monthly price | $50–$100 target | $120/mo Standard |
| Data caps | Not announced | Unlimited Standard; capped Roam tiers |
| Portable tier | Leo Nano — up to 100 Mbps | Roam — $50–$165/mo |
| Business tier | Leo Ultra — up to 1 Gbps | Business — up to 220 Mbps |
| Direct-to-cell (D2D) | Planned 2028 | T-Mobile partnership, live now (texting) |
If Amazon hits the $50–$100/mo target, Leo will be 30–60% cheaper than Starlink's $120/mo Standard plan. Amazon has a Prime distribution channel that Starlink doesn't have, and they don't need to make Leo profitable on its own — it's a loss-leader for AWS uplink demand. Starlink will almost certainly cut prices to compete, so the real win for buyers is the price pressure itself.
Amazon hasn't confirmed a Prime bundle but everyone in the industry assumes it's coming. If Leo Pro shows up at $40/mo for Prime members, that's a fundamentally different product than Starlink — it's the Costco-membership version of internet service. Worth waiting for if you're already a Prime household.
Amazon has shipped tens of millions of consumer hardware units (Echo, Fire TV, Kindle, Ring) and they generally feel polished out of the box. Starlink's terminals work but they look and feel like prosumer-grade hardware. Leo's terminals, if they ship at Amazon's usual polish level, will be a better living-room experience.
Starlink has 6,800+ active satellites; Leo has 331. Starlink has 4+ million subscribers across 100+ countries; Leo has zero. Starlink's onboarding, support, and outage-response workflows have been refined over four years of real residential service. Leo will be the new kid for at least a full year past beta. Whatever bugs Starlink had at launch, Leo will hit too.
Starlink works everywhere on the planet you can see open sky. Leo's beta starts at northern latitudes and expands south — meaning the southern US, Caribbean, and equatorial regions get Leo last. If you live in the South, Starlink may still be your only practical satellite option through most of 2027.
Starlink Standard delivers 50–250 Mbps in real-world use, with peaks above 400 Mbps in low-congestion areas. Leo's 400 Mbps cap is a marketing maximum — actual sustained speeds will be lower depending on how many users are in your cell. Until Leo has a few hundred thousand subscribers under load, the real performance is theoretical.
Starlink's T-Mobile partnership for sat-to-cellphone texting is live in the US already. Amazon Leo's equivalent (Direct-to-Device or D2D) doesn't ship until 2028. If you care about phone-via-satellite as an emergency backup, Starlink has a 2-year head start.
Neither Leo nor Starlink should be compared to traditional geostationary satellite internet (Viasat, Hughesnet). Those services orbit at ~22,000 miles and have 600+ ms latency — fine for email and web browsing, useless for streaming live TV, gaming, or video calls. Both LEO services (Leo and Starlink) are an order of magnitude better. If you're currently on Viasat or Hughesnet, either Starlink today or Leo when it launches will be a massive upgrade.
After 28 years installing residential systems, the truth about satellite internet has always been: it's the option you take when you have no other option. Rural homeowners, off-grid cabins, RVers, people whose fiber-installation quote came back at $80,000 to dig a mile of trench. None of that changes with Leo. Both services need clear sky overhead, both need a small dish on a mounting bracket with line-of-sight to where the satellites pass, and both will rain-fade in heavy weather.
What changes with Leo is the price ceiling on the category. Starlink's $120/mo Standard tier exists because it had no real LEO competitor. Once Amazon enters with a $50–$100 tier and Prime bundle leverage, every satellite price comes down. That helps every rural household whether they end up on Leo, Starlink, or whatever comes next.
If your address has fiber, cable, or a strong 5G fixed-wireless signal — keep what you have. Satellite is still a category for the un-served and under-served, not an upgrade play.