The human race has built about 30,000 spacecraft, ever. There are FCC filings covering more than a million. That gap — roughly 970,000 units — is the market Inflection Space Systems is selling against, and it is a stranger market than the usual launch-deal press release suggests.
The company's first mission is a 36-kilogram cubesat called Pax, booked on an Exolaunch rideshare for no earlier than January 2028. It carries avionics, sensors, actuators, power systems and structures that are meant to fly, later, on a much larger spacecraft.
That larger spacecraft is Exponent, a standardized bus of up to 300 kg at launch, with 100 kg of payload, 1.5 kW of average payload power, 500 m/s of delta-V and a three-year design life. Inflection publishes the price: $1.5 million for one, $700,000 each at a hundred.
The launch is the easy thing to announce. The price list is the actual news, and it is the part most satellite builders keep off the web page.
The bet is manufacturing method — vertical integration and a lights-out factory — not a new spacecraft design.
The near-term risk is qualification and schedule, not demand. The first Exponent has not flown.
The price list is the product
Most satellite primes quote a number after a requirements review, an integration plan and a schedule negotiation. Inflection quotes a table. One bus costs $1.5 million. Order ten and the unit price falls to $1.1 million. Fifty brings it to $900,000. A hundred units, and the hundredth bus is priced at $700,000.
| Units ordered | Unit price (USD) |
|---|---|
| 1 | $1,500,000 |
| 10 | $1,100,000 |
| 50 | $900,000 |
| 100 | $700,000 |
Inflection Space Systems' published Exponent price schedule, 2026.
That slope is the whole argument. A tailored bus spreads its engineering cost across one or two units, so the second spacecraft is barely cheaper than the first. A standardized bus spreads the same engineering across a production run, and the cost per unit falls with volume. The discount is not a sales tactic. It is a statement about how the company intends to make the hardware.
Inflection calls the design "standardized, not modular". The distinction matters. Modular platforms sell optionality and quietly rebuild the integration work on every order. A standardized bus refuses that optionality up front, which is what makes the price curve possible.
The risk in a published price is that it becomes a promise. If the factory cannot hold the curve — if suppliers slip, if a common part needs a redesign, if the qualification campaign stretches — the table turns into a liability the sales team has to explain.
What a standardized bus costs the customer
Standardization is a trade, and the spec sheet states its terms. Exponent fixes payload at 100 kg, average payload power at 1.5 kW, delta-V at 500 m/s and design life at three years. It is launch-vehicle agnostic and stackable. Those limits suit Earth imaging, communications, some sensing and the compute-and-relay satellites now being planned in bulk.
They do not suit everything. A large optical aperture, a high-power radar, a custom-pointing science instrument or a payload that needs a bespoke thermal path will not drop into a fixed envelope. Those missions will keep going to primes that build around the instrument, and they will keep paying for the privilege.
The addressable market is therefore narrower than "spacecraft" and wider than "cubesats". It is the part of the constellation boom that can live inside a standard box. Inflection's demand story — orbital data centers, Golden Dome, direct-to-device — is mostly that kind of mission: many similar units, each unremarkable on its own, none of which justifies a clean-room rebuild.
There is a second-order effect worth naming. If a bus costs $700,000 and a rideshare slot costs a fraction of a launch, the spacecraft stops being the dominant line item in a constellation budget. That changes how operators finance fleets: the constraint moves from "can we afford the satellites" to "can we afford the launches and the ground segment". A cheap bus does not make a constellation cheap. It moves the argument to the next cost.
Why a lights-out factory is the actual moat
Inflection's founders are not pitching a clever mechanism. Alex Lisso came through Rocket Lab, Planet, Millennium Space and Interlune; Matt Hanley through Rocket Lab, Loft Orbital and the Laboratory for Atmospheric and Space Physics. They have watched bespoke spacecraft get built one at a time, and they have concluded the constraint is the production line.
Mass-manufacturing spacecraft requires that we be nearly entirely vertically integrated right away. Pax is an incredible opportunity to get our entire technology stack into orbit at once, gather data, and put those learnings back into our production line.— Alex Lisso, co-founder and CEO, Inflection Space Systems
The company says it has already built more than half of its hardware and software stack in-house, including its own flight software and ground software. It designs for common parts and runs its operations software against simulations, hardware tests and, eventually, orbit. The factory is meant to look more like a fulfillment center than a clean room, with autonomous assembly and round-the-clock lights-out production.
As we wrote in November, the satellite solar array is already turning into a factory product — Gama's €15M round was a bet that a component can be standardized and sold like one. Inflection is pushing the same logic up a level, from a part to the whole bus.
The moat is boring on purpose. It is not a patent. It is a supply chain, a set of tolerances and a willingness to say no to custom work. That is hard to copy, and it is also hard to prove before the first unit exists.
The demand case rests on a number nobody can supply
Inflection's own framing is blunt: the industry has made about 30,000 spacecraft in its entire history, while next-generation constellations — orbital data centers, Golden Dome, lunar infrastructure, direct-to-device connectivity — are being planned in the hundreds of thousands. The company says there are FCC filings for over a million spacecraft. It calls the shortfall a 970,000-spacecraft gap and says it intends to manufacture 10,000 spacecraft a year by 2031.
Total spacecraft humanity has launched
The installed base that next-generation constellations are measured against · Inflection Space Systems / Y Combinator, 2026
The demand is plausible. The hard part is that a filing is not a purchase order. An FCC application is cheap to submit and proves intent, not budget. The companies that file for a million satellites still have to raise the money, win the contracts and clear the regulators before anyone orders a hundred buses.
That is where a standardized bus has an advantage. If demand arrives in bulk, the supplier that can ship in bulk wins. If demand arrives slowly, the same supplier carries an expensive factory with nothing to fill it. Inflection is making an explicit bet that the first case arrives before the factory runs out of runway.
Pax is 36 kilograms of proof — and it is not the product
The pathfinder is deliberately small. Pax is a 16U cubesat, and its job is to earn flight heritage for the hardware and software that will go into Exponent. Inflection frames the mission around two readiness levels: technology readiness, which asks whether the design works, and manufacturing readiness, which asks whether the design can be built in volume. A cubesat can answer the first question cheaply. The second is the one that decides whether the price list survives.
Pax will also carry hosted payloads for two customers. The first is a radiation testbed from The Rad Team and Indiana University's Center for Reliable & Trusted Electronics, built to measure how commercial electronics behave in orbit and to feed real flight data back into radiation models. The second customer has not been named.
The schedule is the standard kind of ambitious. Inflection aims to finish Exponent's critical design and build a first unit by the end of 2027, then ramp production in 2028. The Pax rideshare is set for no earlier than January 2028, which leaves little margin if the first Exponent slips. Exolaunch is providing the launch services; the vehicle and orbit have not been disclosed.
None of that is unusual for a young space company. What is unusual is the public arithmetic. A buyer can now compare a bespoke bus quote against a published curve and see the gap. That pressure runs both ways: it disciplines the incumbents, and it gives Inflection's customers a number to hold the company to.
The interesting thing about Inflection is not that it will fly in 2028. Plenty of companies will. The interesting thing is that it has already told you what the hundredth unit costs, and invited you to check.