SpaceX Plans Reported $8 Billion Spectrum Purchase to Build Starlink Mobile Into a US Carrier

A nationwide spectrum deal brings Starlink closer to the ground

SpaceX agreed on October 8, 2026, to acquire Grain Management's nationwide 800 MHz spectrum portfolio, adding airwaves that could help Starlink Mobile reach phones inside buildings. The proposed purchase covers up to 14 MHz of paired spectrum and licenses serving approximately 100% of the US population, although the amount available varies by county.

The companies did not disclose financial terms. Reports citing people familiar with the transaction put the price at about $8 billion in cash. The agreement still requires Federal Communications Commission (FCC) approval and other closing conditions, so it does not yet give SpaceX an operating nationwide cellular network.

The purchase follows authorization for a separate constellation of 15,000 Starlink Mobile satellites and a regulatory waiver that reduces SpaceX's dependence on spectrum agreements with established carriers. Together, those steps give the company a clearer route toward selling mobile service under its own control, rather than primarily supplying satellite coverage to another operator.

SpaceX CEO Elon Musk described the acquisition as the final spectrum component needed for his company's coverage ambitions. In a social media post on Thursday, he said:

This is the last critical piece of the spectrum puzzle needed for SpaceX to provide complete phone coverage in America.

That is an ambition, not a demonstrated service result. The central distinction is between obtaining frequencies that travel farther and through obstacles, and building a network with enough capacity to serve millions of people simultaneously. SpaceX's plans still require ground equipment, satellite deployment and commercial decisions that have not been announced.

What SpaceX is buying

The portfolio includes frequencies from 817 to 824 MHz paired with 862 to 869 MHz. Paired spectrum provides separate channels for traffic traveling from a phone to the network and back again. The maximum 14 MHz describes the combined width of those channels, not a guaranteed connection speed.

An FCC order approving Grain's earlier acquisition described the licenses as covering approximately the entire US population, with between 4.85 MHz and 14 MHz available depending on the county. Geographic reach is therefore extensive, but the quantity of spectrum is neither uniform nor especially large.

The airwaves came from T-Mobile, which obtained them through its 2020 acquisition of Sprint and was required to sell them under a government merger condition. The licenses trace back to Nextel. Dish had an initial option to buy them for $3.59 billion, but declined, and that option expired in 2024.

Grain's purchase was announced in early 2025, received FCC approval in July 2026 and closed in August. Grain paid $2.9 billion in cash and transferred 600 MHz licenses covering 10 MHz in 212 counties, representing approximately 15% of the US population. Comparing that cash payment directly with SpaceX's reported $8 billion price would omit the value of the spectrum Grain also contributed.

New Street Research analyst David Barden valued the reported SpaceX transaction at about $1.73 per MHz per person covered, above a speculated Grain target of $1.30. This measure compares spectrum prices by adjusting for both the amount of bandwidth and the population within the licensed territory. It is an analyst calculation based on a price the companies have not confirmed.

Why indoor coverage still needs ground equipment

Lower frequency radio signals generally travel farther and pass through walls and vegetation more effectively than higher frequency signals. SpaceX says the 800 MHz licenses will form a coverage layer alongside its 2 GHz spectrum, which is intended to provide greater capacity. Most existing mobile devices already support the 800 MHz band, according to the company.

That does not establish that existing Starlink satellites can immediately transmit on these frequencies. One account specifically says SpaceX's current V1 and planned V2 mobile satellites are optimized for the EchoStar 2 GHz spectrum and are not currently compatible with the acquired 800 MHz band. The purchase is therefore best understood as supporting a combined satellite and terrestrial network, rather than simply switching on a new satellite frequency.

SpaceX's FCC filing describes radios feeding antennas on towers, rooftops and other structures. Musk has also discussed using many small base stations and connecting them through Starlink broadband equipment on homes and businesses. Reports describe a role for femtocells, small cellular base stations that provide local coverage.

These approaches need not be mutually exclusive. Small cells could serve particular buildings or neighborhoods, while leased towers or other larger installations provide broader coverage. Reliable urban service would also require connections that carry traffic from those radio sites into the wider network, through fiber or other suitable links.

Tim Farrar of TMF Associates said the acquisition intensifies competition, but warned that reliable reception inside urban buildings would require towers on the ground. The practical issue is not just whether a signal reaches a phone. It is whether the network can sustain calls and data traffic when many nearby customers connect at once.

The satellite approval is a separate milestone

The FCC has authorized SpaceX to deploy and operate up to 15,000 satellites for its new mobile system, which was proposed in September 2025. The approved orbits are between 326 and 335 kilometers above Earth. Flying closer to users can reduce transmission delay, but authorization is not equivalent to deployment.

Reports differ on the approval date. Two give October 6, 2026, while another gives October 7. They agree on the 15,000 satellite authorization and on a waiver allowing SpaceX to provide satellite wireless service without the usual spectrum leasing arrangement with a terrestrial mobile operator.

The new mobile constellation is separate from the Gen2 broadband system. In January, the FCC authorized another 7,500 Gen2 satellites, bringing that system's permitted size to 15,000. The two authorizations therefore cover 30,000 satellites across the broadband Gen2 and new mobile systems combined, excluding the separately authorized first generation constellation.

For comparison, tracking data from astronomer Jonathan McDowell showed nearly 13,000 Starlink satellites launched by early October, with around 11,150 still in orbit. Those figures describe launches and spacecraft remaining in orbit, not 13,000 satellites already delivering the newly approved mobile service.

SpaceX says it has launched approximately 650 satellites for its current mobile offering. It claims the V2 satellites will deliver more than 100 times the bandwidth of the current generation, with speeds reaching up to 150 Mbps per user. Current T-Satellite speeds are estimated at around 4 Mbps. These are different measures: total satellite bandwidth describes shared capacity, while a user speed depends on coverage, demand and operating conditions.

EchoStar spectrum and the deployment calendar

The Grain purchase complements a separate agreement to acquire 65 MHz of nationwide spectrum from EchoStar for about $17 billion. That transaction was agreed in September 2025 and approved by the FCC in May 2026. The mobile satellite authorization covers five EchoStar frequency bands, alongside the T-Mobile frequencies used for the existing service.

The EchoStar bands run from 1695 to 1710 MHz, 1915 to 1920 MHz, 1995 to 2000 MHz, 2000 to 2020 MHz and 2180 to 2200 MHz. The T-Mobile authorization covers 1910 to 1915 MHz and 1990 to 1995 MHz. These approvals give SpaceX access to several frequency blocks, rather than a single uninterrupted channel.

SpaceX can only test on the EchoStar frequencies until the second stage of that acquisition closes, scheduled for November 30, 2027. The FCC approval included a $2.4 billion escrow condition. Decisions on requested use of 2020 to 2025 MHz in the United States and 1429 to 2690 MHz outside the United States were deferred.

The sequence of regulatory and commercial milestones matters because different parts of the proposed network become usable at different times:

• September 2025: SpaceX submitted the new mobile constellation application and agreed to the EchoStar spectrum purchase.
• May 2026: The FCC approved the EchoStar spectrum transfer.
• July and August 2026: Grain received approval for, then completed, its purchase of T-Mobile's 800 MHz licenses.
• October 6 or 7, 2026: Reports place the new mobile satellite authorization on these differing dates.
• October 8, 2026: SpaceX and Grain announced their proposed spectrum transaction.
• November 30, 2027: The second stage of the EchoStar transaction is scheduled to close.

SpaceX previously told the Canadian government that service from its next generation mobile satellites would begin by late 2027, with thousands more satellites launched by the end of 2028. Japanese carrier KDDI also plans service outside terrestrial coverage by the end of 2027, including native voice calls, streaming, browsing and more demanding applications. Neither timetable establishes a launch date for an independent nationwide US retail service.

Coverage is not the same as capacity

T-Mobile Chief Technology Officer John Saw says Starlink Mobile would have roughly 80 MHz of spectrum after the acquisition, some usable only for transmissions toward the network. He compares that with T-Mobile's nearly 400 MHz of medium and low frequency holdings. On those stated figures, T-Mobile has about five times as much spectrum, although the bands and permitted uses differ.

Saw said the US wireless industry has invested more than $250 billion in network infrastructure since 2018, alongside more than $200 billion in spectrum auctions during the past decade. His argument also covers engineering, site acquisition, customer support and distribution, all of which SpaceX would need to address as a retail operator.

In his response to the deal, Saw drew a distinction between buying licenses and delivering service:

You can acquire spectrum in a single transaction. You cannot acquire a nationwide, high-performing wireless network overnight.

Analyst Vish Nandlall similarly argues that the purchase improves coverage more than capacity. A limited spectrum block can extend service into rural areas or buildings, but busy urban networks need enough bandwidth and radio sites to divide traffic among many users.

Cost estimates vary widely. Bernstein Research puts a competitive national network at $50 billion to $130 billion. BNP Paribas previously estimated $100 billion and 50,000 to 70,000 sites, potentially rising to 120,000 with additional C band holdings. TD Cowen analyst Gregory Williams estimates that construction could take years and cost upwards of $80 billion. These are analyst scenarios, not SpaceX construction budgets.

Using Nandlall's example, recovering $50 billion over ten years from 20 million subscribers would require approximately $21 per subscriber each month, before operating costs, financing or a return on investment. That calculation helps explain why buying wholesale cellular capacity or leasing existing infrastructure could be more practical than duplicating every part of the established networks.

Partners are preparing for competition

SpaceX currently provides US mobile satellite service through T-Mobile's T-Satellite, launched in July 2025, using 10 MHz of leased nationwide spectrum. The partnership remains intact, even though T-Mobile has removed Starlink's name from its main T-Satellite website.

AT&T, T-Mobile and Verizon have also created a joint venture to expand satellite coverage in areas their terrestrial networks do not reach. The initiative was announced in May and formally launched on October 1. Those dates explain accounts that describe the venture as either a May development or a recent agreement.

The carriers say existing satellite agreements remain in place and that they can pursue connectivity projects independently. Their venture seeks broad device compatibility through work with operating system providers, app developers and equipment manufacturers. Starlink is not part of it.

LightShed Partners analysts Walter Piecyk and Joe Galone questioned how much spectrum the venture would actually receive. They also warned that coordination intended to exclude SpaceX could raise antitrust concerns. That is an analyst warning, not a finding of unlawful conduct.

Carrier relationships could still shape SpaceX's entry. Williams has pointed to a mobile virtual network operator arrangement, in which SpaceX would sell service using wholesale access to another operator's network, as an easier alternative to nationwide construction. Its KDDI agreement also shows that gaining greater independence in the United States does not mean abandoning partnerships elsewhere.

Investors react before a service arrives

The announcement drove sharp declines in carrier shares, although the figures differ by trading session. One account recorded T-Mobile falling more than 11% after hours, with AT&T and Verizon down about 6.6%. Another described declines of around 6% across the three in extended trading.

Friday morning figures put T-Mobile down $18.58, or 11%, at $152.72, and AT&T down $1.83, or 7.4%, at $22.76. SpaceX rose $2.61, or 1.6%, to $163.18 in that account. Other snapshots put SpaceX's gain around 2% or 2.5%. These are observations from different trading windows, not one common closing result.

Tower stocks moved in the opposite direction. One early trading account put Crown Castle up 12%, American Tower up 7% and SBA Communications up 6%. The contrast reflects two possible outcomes being priced at once: more competition for mobile customers and more demand for the ground infrastructure SpaceX may need.

UBS analyst John Hodulik expects existing handset compatibility to help and does not foresee regulatory obstacles. He nevertheless expects scaling to take time, with limited immediate effects on wireless industry finances. Morningstar analyst Michael Hodel raised his uncertainty rating for T-Mobile, arguing that even the threat of SpaceX entering could disrupt the current market structure.

A TD Cowen survey found that 34% of US consumers and 30% of T-Mobile customers would consider switching to Starlink Mobile. Interest is not a subscriber forecast, particularly before pricing, availability and actual performance are known.

What remains unresolved

Starlink gives SpaceX an existing customer base and a profitable connectivity business. Its second quarter figures showed 12 million subscribers as of June 30, availability in 205 countries, revenue of $4.29 billion and operating income of $1.66 billion. Deutsche Bank estimated 13.5 million subscribers worldwide at September's end, including 4.1 million in the United States. The September total is an analyst estimate, not the same kind of company disclosure as the June count.

Those customers could provide a market for broadband and mobile bundles, but they do not establish demand for a separate cellular subscription. SpaceX has not announced retail pricing, a firm US launch date, tower agreements or nationwide indoor performance results.

The Grain transfer also faces regulatory scrutiny. The Rural Wireless Association opposed earlier EchoStar spectrum sales, saying concentration disadvantages rural providers. That objection does not establish that the Grain transaction will be blocked. FCC Chair Brendan Carr has spoken favorably about the investment and consumer opportunities the deal could bring.

Further spectrum decisions are approaching. The FCC planned an October 29 vote on seeking public comment about making another 482 MHz available for satellite coverage and updating its rules. It was also considering an auction of 25 MHz for satellite and terrestrial services. Neither proposal is spectrum already awarded to SpaceX.

Deployment brings separate concerns. The American Astronomical Society raised issues involving interference with observations, orbital debris and atmospheric effects from satellite reentries. SpaceX must still launch the authorized constellation and demonstrate the promised service. The deal strengthens its position, but spectrum ownership alone does not settle network quality, construction costs or commercial viability.

Key Points

• SpaceX agreed to buy Grain Management's nationwide 800 MHz portfolio, with up to 14 MHz of paired spectrum.
• The reported price is about $8 billion in cash. The companies have not disclosed terms, and FCC approval is still required.
• The frequencies could improve terrestrial coverage inside buildings, but ground infrastructure remains necessary.
• A separate FCC authorization permits 15,000 new mobile satellites. Reports differ between October 6 and October 7 on the approval date.
• The second stage of the EchoStar spectrum purchase is scheduled to close on November 30, 2027.
• Carrier shares fell, while tower companies gained as investors weighed competition and infrastructure demand.
• SpaceX has not announced pricing or a firm launch date for an independent nationwide US mobile service.