SpaceX is gearing up to turn the sky into a massive cellular network. In a new filing, the company asked regulators to approve a next-generation constellation of up to 15,000 direct-to-cell “V3” satellites designed to beam 5G service straight to phones on the ground. SpaceX says these V3 satellites will deliver up to 20 times the throughput of today’s direct-to-cell V2 units, marking a major leap in satellite-to-phone connectivity.
Today’s Starlink direct-to-cell tests are limited to messaging and small amounts of data in optimized apps through partnerships such as T-Mobile. With the V3 network, SpaceX aims to go beyond basic messaging and enable real 5G data directly to standard handsets. The plan includes expanding Starlink’s extensive network of ground stations to boost coverage, increase speeds, and cut latency. According to the filing, Starlink’s satellite-to-phone service could reach up to 100 Mbps on supported phones using a dedicated satellite modem chip, roughly matching the performance you can get from a Starlink Mini dish—without any external hardware.
SpaceX’s ambitions don’t stop at wholesale connectivity. While the company is working with carriers to provide satellite coverage, it is also positioning itself to operate as a space-based carrier in its own right. Following a sizable spectrum deal reportedly valued at $17 billion with the parent company of Dish, Elon Musk said users “should be able to have Starlink like you have an AT&T, or T-Mobile, or Verizon” in the future. Even if traditional carriers still “own a lot of spectrum,” a 5G satellite link delivering 100 Mbps would be a real milestone—on par with the median LTE speeds major networks were posting just last year.
The scale of the hardware required is staggering. The V3 satellites are designed to be immense—roughly the size of a Boeing 737 when their arrays are fully unfurled. SpaceX plans to launch them using the Starship 3 rocket, with a maiden flight targeted for this year. If all goes to plan, V3 satellite launches could begin in 2026, coinciding with the activation of a gigabit-class Starlink network for dedicated equipment such as the new Performance dish. That high-throughput layer would complement the direct-to-cell constellation focused on phone-grade connectivity.
Why this matters is simple: universal coverage. A direct-to-phone 5G network from space could finally erase dead zones, deliver service to remote communities, improve connectivity at sea and in the air, and provide critical backup during disasters when terrestrial networks fail. The approach also promises lower latency and more consistent speeds as SpaceX builds out additional ground gateways and optimizes routing across its constellation.
Key details at a glance:
– Up to 15,000 V3 direct-to-cell satellites requested in a new FCC filing
– V3 aims for 20x the throughput of current V2 direct-to-cell units
– Target speeds up to 100 Mbps to compatible phones with a satellite modem chip
– Expanded ground infrastructure to improve coverage and latency
– Partnerships with carriers continue, while SpaceX pursues its own space-based carrier model
– Launches depend on successful Starship 3 testing; V3 deployments targeted for 2026
– Gigabit service remains available for dedicated gear like the Performance dish
There are still big hurdles. The entire plan hinges on Starship 3 proving reliable enough to lift and deploy these massive satellites at scale. Regulatory approvals, spectrum coordination, handset support, and seamless carrier integration are all critical pieces that must fall into place. But if SpaceX executes, the result could be a step change for mobile connectivity: true 5G data from orbit to ordinary phones, and a new competitive dynamic in the wireless industry.
For consumers, the immediate takeaway is that satellite-to-phone service is evolving from basic texting into something much closer to everyday mobile data. For carriers and competitors, the message is clear: space-based networks are moving quickly from concept to coverage. And for the vast parts of the world where terrestrial towers are sparse or nonexistent, that could be the most transformative development of all.
