ASTS — Q2 2026 Financial Report Analysis

Revenue arrives, but a lost satellite and a slipped launch target set the tone

AST SpaceMobile is building a network of very large satellites in low Earth orbit that talk directly to ordinary, unmodified smartphones — "direct-to-device" service, so a phone with no signal from a cell tower can still connect through space. It sells that capacity through mobile network operators (MNOs) such as AT&T, Verizon and Vodafone rather than to consumers. The paid service itself has not launched yet; today's revenue comes from selling ground equipment to those operators and from U.S. government contracts.

In the second quarter of 2026 (April–June), revenue was $31.5 million, up from $1.2 million a year earlier and $14.7 million in Q1. The net loss attributable to shareholders widened to $230.9 million ($0.77 per share) from $99.4 million ($0.41). More than half of the jump came from one item: a $125.9 million write-off of the BlueBird 7 satellite, which a launch vehicle placed into too low an orbit on April 19 and which then had to be brought down. The quarterly filing also moves the company's target of roughly 45 satellites in orbit from "by the end of 2026" (Q1 filing) to "early 2027".

At a glance

• $31.5 million revenue — real money, but still only 31% of the bottom of management's $150–200 million full-year guidance after six months ($46.3 million so far), so second-half revenue has to be roughly 2.2 to 3.3 times the first half's.
• $97 million of operating cash burn in Q2 (H1: $145.2 million), plus $859 million spent on satellites and equipment in H1 — the business consumed about $1 billion of cash in six months before any service revenue.
• $2.7 billion of cash (including $434.6 million restricted) at June 30, and over $3.7 billion pro forma after a $1.15 billion convertible bond sold in July — management says it is fully funded for about 90 satellites.

The numbers

n/m = not meaningful. Percentage changes on a revenue base of about $1 million, or between two losses, would produce eye-catching numbers that say nothing useful, so we leave them out.

Where the revenue came from

Products revenue of $24.4 million was, per the filing, "primarily attributable to sales of gateway equipment and software to MNOs" — gateways are the ground stations that connect the satellites to an operator's network. That business is close to a pass-through: cost of product revenues was $22.4 million, leaving about $2.0 million (an 8% margin). $1.9 million of the product sales went to SatCo, the European joint venture AST owns 50/50 with Vodafone, so a slice of revenue is effectively sold to itself and its partner.

Services revenue of $7.1 million came from completing milestones on U.S. government contracts, at a much higher margin ($1.2 million of cost). The press release says AST received U.S. government awards worth over $125 million in aggregate and that contracted revenue backlog — revenue already signed but not yet earned — rose to about $1.30 billion.

Where the costs went

Leaving aside the satellite write-off, the cost base is growing fast and it is mostly people:

• Engineering services rose $58.7 million to $87.3 million, of which $34.7 million was "payroll and employee related costs driven by an increase in headcount and higher stock-based compensation expenses", plus $12.2 million more in consulting fees.
• General and administrative costs rose $36.7 million to $63.9 million, $30.6 million of it payroll and stock-based pay.
• Depreciation rose 76% to $20.7 million, from lab, assembly and integration equipment.

Management's preferred measure, "adjusted operating expenses" (which removes stock-based pay, depreciation and the BB7 loss), was $119.1 million in Q2 versus $91.2 million in Q1.

What the headline numbers hide

The BB7 loss is a genuine one-off, but satellite risk is not. The $125.9 million is the write-off of BlueBird 7 net of insurance; the company has collected $21.6 million of insurance and booked another $10.9 million as a receivable, and expects a free replacement launch under its launch contract. Strip the write-off out and the loss before minority interests was about $174 million (our arithmetic: $299.9 million minus $125.9 million). The comparable year-ago figure, stripping a $65.0 million warrant-revaluation loss, was about $71 million — so the underlying loss still roughly 2.5x'd.

Stock-based pay is now a large part of the expense line. Stock-based compensation was $63.5 million in the quarter, six times last year's $10.5 million, and the filing shows another $289.8 million of unrecognized cost from the 2024 plan to be expensed over about 2.4 years. It doesn't cost cash, but it is paid in shares, which dilutes existing holders.

Cash conversion is the real story, and it is negative by design. Operating cash outflow was $97 million in Q2 (the six-month figure of $145.2 million less Q1's $48.1 million), smaller than the accounting loss because of non-cash charges. The bigger outflow is building the constellation (a "constellation" is the fleet of satellites that together give continuous coverage): $859.2 million of property and equipment in six months, plus a $100 million advance tied to the Ligado spectrum deal (now held in escrow by a bankruptcy court order) and $42.1 million of spectrum payments.

Receivables jumped. Accounts receivable rose to $79.3 million from $37.7 million at December 31, more than the $46.3 million of revenue booked in the whole first half. Some of this is timing on milestone billings, but it is worth watching that customers pay. Separately, a $45.0 million prepayment from Verizon became due after AST received regulatory approvals on April 22, 2026, and per the 10-Q had not yet been received as of the filing date.

Dilution has been the price of the funding. Weighted-average Class A shares rose 23.6% year on year to 299.1 million. In February the company repurchased $296.5 million face value of convertible notes by issuing about 6.3 million new shares, booking an $89.8 million "induced conversion" charge in Q1 for the extra value it paid holders to convert early. Total debt (carrying value) rose to about $2.97 billion from $2.22 billion at year-end, almost all low-coupon convertible notes, which can turn into shares later; July's $1.15 billion 1.625% note adds to that. Interest expense ($26.1 million) was covered by interest earned on the cash pile ($27.7 million) this quarter.

Guidance now vs before. The $150–200 million 2026 revenue target is unchanged from Q1, when management said about half was already covered by contracted backlog. But the satellite deployment target has moved: the Q1 10-Q said the company would "continue to target approximately 45 BB satellites by the end of 2026"; the Q2 10-Q says it is "targeting approximately 45 BB satellites in early 2027." Average capital cost per satellite is still estimated at $21–23 million.

Takeaway: The revenue line is finally moving, but it is gateway hardware sold at an 8% margin and government milestones, not the satellite service the company is valued on. What decides the next year is satellite count — 25 BlueBirds are needed for limited service and 45–60 for continuous coverage of key markets — and the Q2 filing quietly pushed the 45-satellite target from end-2026 into early 2027 after losing BlueBird 7.

Where the network stands

Counting from the filing: five first-generation Block 1 satellites (launched September 2024) plus seven larger Block 2 satellites in orbit — BB6 (launched December 2025), BB8–BB10 (June 2026) and BB11–BB13 (August 5, 2026) — so 12 BlueBirds, with BB7 lost. The press release counts 13 spacecraft in orbit overall. Each Block 2 satellite carries an antenna of up to about 2,400 square feet, which the company says is the largest commercial array in low Earth orbit and is what lets a weak signal from a normal phone be heard from space.

Management's own thresholds, from the 10-Q:

• 25 satellites (5 Block 1 + 20 Block 2): limited, "noncontinuous" service in targeted markets — meaning coverage with gaps during the day.
• 45–60 satellites: continuous service across the U.S., Europe, Japan and other key markets.
• About 90 satellites: continuous service in all targeted markets; the company says it is fully funded to build and launch these.

The release says BB14–BB16 are about to ship and BB17–BB46 are in production, with factory capacity built for up to six satellites a month. It also mentions about 3,000 "digital cells" activated across the continental U.S. ahead of a beta service in 2026 with selected operators. On spectrum, the Ligado deal (up to 45 MHz of mid-band spectrum in the U.S. and Canada) still awaits regulatory approval and is entangled in Inmarsat's appeal of a bankruptcy court ruling.

Outlook

Management's guidance: 2026 revenue of $150–200 million, said to be "on track"; beta service with select partners in 2026; about 45 satellites in orbit in early 2027; average cost of $21–23 million per Block 2 satellite.

Our read: The revenue guidance needs about $104–154 million in the second half, or roughly $52–77 million a quarter against $31.5 million in Q2. That is reachable only if gateway shipments and government milestones keep accelerating, and gateway sales add little profit even when they arrive. The more important deadline is physical: with 12 BlueBirds up and three more about to ship, reaching the 25 needed for initial service depends on launch cadence over the next few months, and the company itself has just added a quarter or so to its 45-satellite target. The balance sheet is not the constraint near term — over $3.7 billion pro forma covers years at the current burn — but the funding has come through steady share issuance and convertible debt, so per-share economics keep diluting until service revenue shows up. Things to watch in Q3: the satellite count at quarter-end, whether the Verizon $45 million prepayment and the higher receivables are collected, any change to the revenue range, and whether the FCC approves the Ligado spectrum transfer.

This is AST SpaceMobile's first report on this site, so there is no earlier outlook to check against.