In July 2026, NASA announced it was committing nearly $600 million (a combined $590.4 million) to four Moon landings planned for late 2028. The deliveries will be handled not by a government agency but by three private companies — Astrobotic, Firefly Aerospace, and Intuitive Machines [source: NASA, 2026]. The intriguing part is that all four landers will carry the same set of science instruments: a trio built to measure lunar navigation, the dust kicked up during landing, and the radiation environment.
The announcement means more than a line in a budget. It signals that a model is taking firmer hold — one in which NASA, rather than building and flying its own landers, pays a "delivery fee" to privately built landers to ferry science instruments to the Moon. That model is Commercial Lunar Payload Services (CLPS). This article separates three things carefully: what this contract has confirmed (the money, the companies, the schedule, the instruments), the actual track record (how the landings have gone so far), and what remains strategic intent still being tested (the promises and risks of a low-cost commercial model).
A word on method, because those layers carry very different weight. Some things are confirmed by contract: the money, the companies, the instruments NASA is buying rides for. Some are measured: how the attempted landings actually ended, and what a government audit found when it added up cost and schedule. And some are intent: the strategy behind a low-cost commercial model, a plan under test rather than a result. One note on dating — NASA's release and the outlets that republished it place the announcement anywhere from late June to mid-July 2026, so this article says "July 2026" without pinning a day.
How this article is organized
- What NASA actually announced
- Why fly the same instruments four times
- A new model: NASA buys the ride, not the lander
- How the landings have gone so far
- The criticism and the risks
- Conclusion — what to watch
What NASA actually announced
The confirmed numbers
Start with the confirmed figures. NASA handed three companies a total of four lunar deliveries. Astrobotic gets $297.9 million for two, Firefly Aerospace $144.2 million for one, and Intuitive Machines $148.3 million for one [source: NASA, 2026]. Together that is about $590 million, which NASA rounded to "nearly $600 million."
The split shows what one lunar delivery currently costs. Firefly's award covers one delivery for $144.2 million and Intuitive Machines' one for $148.3 million, while Astrobotic's larger $297.9 million buys two [source: NASA, 2026]. These are prices in the $100-million range — the band that has characterized CLPS awards from the beginning [source: New Space Economy, 2025]. And what the money buys is not hardware NASA will own: a delivery, bought four times over.
The schedule, the program, and one caveat
All four landings target late 2028. The missions are part of NASA's Moon Base Program and its Artemis campaign, and they carry additional science and technology-demonstration payloads to the lunar surface. Lori Glaze, NASA's associate administrator for the Human Spaceflight Mission Directorate, said the "new awards demonstrate our commitment to accelerating our effort to build a long-term presence on the lunar surface" [source: NASA, 2026]. Keep in mind, though, that "late 2028" is a target date — and as we will see, CLPS missions have a history of slipping.
Be precise about what kind of date this is. "Late 2028" is a target attached to a contract, not a measured event, and this program carries a documented record of delay — an average of more than 14 months per task order, by NASA's own inspector general's count [source: NASA OIG, 2024]. That does not make the date meaningless — only a projection best read alongside the program's own record.
It helps to know what these deliveries are for. Beyond the science they carry, CLPS missions are meant to demonstrate systems ahead of crewed Artemis missions, lay groundwork for a long-term human presence, and grow a commercial lunar market [source: NASA, 2026]. The three instruments riding on these landers — navigation, dust, radiation — line up with that first purpose. They are reconnaissance for the landings meant to carry people.
Why fly the same instruments four times
The trio, and what each one is for
The most striking feature of this contract is that all four landers carry the same three instruments. The first is the LRA (Laser Retroreflector Array). With eight quartz prisms that bounce back a laser fired from an orbiter, it lets spacecraft fix their position precisely — a passive navigation marker that needs no power and no maintenance. The second is SCALPSS (Stereo Camera for Lunar Plume Surface Studies), which uses four cameras in stereo to record in three dimensions how a landing engine's exhaust scatters lunar dust, improving models that predict dust erosion. The third is LETS (Linear Energy Transfer Spectrometer), a compact silicon detector that measures the space-radiation environment at each landing site to assess the exposure risk to astronauts [source: NASA, 2026].
Each is deliberately modest, and each answers a question a permanent lunar outpost would need answered. The LRA's eight prisms are corner-cube reflectors — a geometry that returns a beam along the path it arrived on — which is why a passive tile with no electronics works as a survey marker [source: NASA, 2026]. Needing neither power nor upkeep, its usefulness outlasts the lander that carried it. SCALPSS asks what landing does to the ground underneath — plume-thrown dust is one of the hazards NASA says it wants to understand better. LETS asks the bluntest question: how much radiation would a person standing on that spot absorb?
Why standardization is the point
The reason for flying the same kit over and over is simple. Joel Kearns, NASA's deputy associate administrator for exploration in the Science Mission Directorate, explains that "by flying the same science instruments on multiple landers, we will better understand potential hazards during landing and build out a global network of environmental data and location markers on the Moon" [source: NASA, 2026]. Different instruments at each site would be hard to compare against one another; a standardized kit spread across many sites lets NASA gather dust, radiation, and position data from across the Moon on a single yardstick. The aim is not a single landing but a network of points.
The comparison problem is the practical heart of it. Instruments differ in calibration, sensitivity, and mounting, so two different radiation detectors at two sites return readings awkward to place on one scale — you never know how much of the gap belongs to the Moon and how much to the hardware. Flying one design repeatedly removes that ambiguity: differences between sites can be attributed to the sites. That is what turns four landings into one dataset.
From four landings to a network
The framing also explains why the number of landings that actually arrive matters. A marker network gains value with every added point; four identical kits are a plan for four points, and how many become real points is a separate question. That is what makes the landing record in section 4 more than a scoreboard — it decides how much of this science gets built.
A new model: NASA buys the ride, not the lander
What CLPS actually is
The framework holding all of this up is CLPS. It runs on indefinite-delivery, indefinite-quantity (IDIQ) contracts with a combined maximum value of $2.6 billion through November 2028 [source: NASA, 2026]. The key is the nature of the contract. Each award bundles payload integration, mission operations, launch from Earth, and lunar landing into a single end-to-end delivery service. In other words, NASA does not own the lander — it buys a seat, a ride, aboard one. Established in 2018, the program now keeps a pool of 13 eligible American companies and has so far awarded 17 deliveries to five vendors to carry more than 60 instruments to the Moon [source: NASA, 2026].
Two pieces of contract jargon do a lot of work there. "Indefinite-delivery, indefinite-quantity" means the agreement fixes the terms and a ceiling value but not how many deliveries NASA will order or when; individual missions are placed against it as task orders [source: NASA, 2026]. "End-to-end" means the vendor owns the whole chain, from integrating payloads to setting the lander down. In a traditional program the agency would be managing a spacecraft. Here it manages a purchase.
Awarded versus flown
Hold two of those figures next to each other. CLPS has awarded 17 deliveries to five vendors and keeps 13 companies eligible, yet through early 2026 only four of those deliveries had attempted a landing [source: NASA, 2026]. The gap is no scandal — landers take years to build — but most of this program still exists as contracts, not flight history. The four awards announced in July 2026 add to the contracted column, not the flown one.
The "shots on goal" strategy
Underneath the model sits a philosophy often called "shots on goal": let many companies try in parallel, quickly and repeatedly, to gain speed. The crucial point is that NASA does not expect every attempt to succeed. Because each contract is relatively cheap — generally in the $100 million range — NASA is willing to accept a success rate of perhaps only about half in exchange for moving fast [source: New Space Economy, 2025]. The bet is that many inexpensive tries beat one expensive certainty. But that "accepted risk" shows up plainly in the actual landing record (see section 4).
The logic only holds because of the price tag. A program that expects to lose roughly half its attempts would be hard to defend if each were a flagship-sized investment; it becomes arguable when each contract sits in the $100 million range and several run at once [source: New Space Economy, 2025]. The trade is explicit: lower odds per mission in return for more missions, sooner, from more suppliers. Whether it pays off is settled by how the attempts land.
The bigger picture: a lunar economy
CLPS's larger goal reaches beyond science to growing a commercial lunar market itself: to seed a sustainable lunar economy and reduce NASA's dependence on government-built spacecraft. In this design, NASA is less a monopolist than an anchor customer that adds to demand — priming the pump so private landers can carry cargo for others too and a market can eventually stand on its own [source: Payload, 2025]. As demand has grown, NASA moved in April 2026 to raise the CLPS contract ceiling from $2.6 billion to $4.2 billion [source: SpaceNews, 2026].
The anchor-customer idea is the load-bearing assumption of the model. An anchor customer buys enough, reliably enough, that a supplier can justify capacity it would not otherwise build — but the point is that other buyers eventually join it. If that second wave never appears, CLPS does not produce a lunar economy; it produces contractors whose only customer is NASA. Nothing in the contracts settles that. It is the part of the model only time, and other people's money, can test.
The demand signal from NASA's own side, at least, points upward. The April 2026 move to lift the ceiling came from NASA's Johnson Space Center, which framed it as a response to a surge in lunar lander missions [source: SpaceNews, 2026]. Note the status of that number: raising a ceiling authorizes spending, it does not commit it. A $4.2 billion ceiling is headroom for orders NASA may place — not $4.2 billion of awarded work.
How the landings have gone so far
The first two attempts
So much for the strategy — how has it actually played out? Through early 2026 there had been four CLPS landing attempts, and the scorecard deserves a clear-eyed look. The first, Astrobotic's Peregrine (January 2024), never reached the Moon at all: a propulsion anomaly struck hours after launch [source: NASASpaceFlight, 2025]. The second, Intuitive Machines' IM-1 "Odysseus" (February 2024), reached the surface but tipped onto its side. It still held on for about a week and salvaged some data [source: NASASpaceFlight, 2025].
Those two outcomes show that lunar failure is rarely binary. Peregrine never got to attempt a landing — its trouble arrived hours after launch, long before the Moon. Odysseus did land, then spent about a week on its side returning some of the data it was sent to collect [source: NASASpaceFlight, 2025]. Scoring either as a plain zero throws away the record's most useful information: where in the chain — launch, cruise, descent, touchdown — the difficulty actually sits.
The breakthrough, and the setback four days later
The turning point was the third. Firefly Aerospace's Blue Ghost Mission 1 (March 2, 2025) touched down fully successfully and upright — the first fully successful landing by a U.S. commercial lander since 1972, which NASA called "proof positive" of CLPS [source: SpaceNews, 2025]. Yet four days later, on March 6, Intuitive Machines' IM-2 "Athena" tipped over again and met only some of its objectives [source: NASASpaceFlight, 2025]. In sum, one of the four attempts was a full success. Rather than reading that record as "the commercial model failing," it is more accurate to read it as hard measured evidence of just how difficult a soft lunar landing is on the first serious return in half a century.
Reading the scorecard honestly
It is worth resisting the tidy conclusion that a one-in-four record indicts commercial procurement. Soft-landing on the Moon is genuinely difficult, and these attempts came on the first serious American return to the surface in more than half a century [source: SpaceNews, 2025]. Whether a government-run program attempting the same four landings would have done better is unknowable from so small a sample. What the record does establish is that NASA's willingness to accept roughly even odds was a forecast — and so far the flights have matched it.
The criticism and the risks
What the audit measured
The picture is not all optimism. NASA's Office of Inspector General (OIG) spelled out CLPS's downsides in a 2024 report, IG-24-013. Across the missions there was a $208.2 million cost increase, and each task order slipped by an average of at least 14 months. A schedule that had assumed an average of 30 months from award to launch actually ran 44 months — roughly 50% longer [source: NASA OIG, 2024]. The auditors blamed "overly optimistic market research" that underestimated supply-chain and technical difficulty, and found that folding in the expensive VIPER rover ($433.5 million) early on added to the cost and schedule pressure.
Two of those findings compound each other. A task order that runs 44 months instead of 30 is not merely late; under a fixed-price contract it is 14 extra months in which a small vendor carries its overhead against a payment that does not grow [source: NASA OIG, 2024]. The auditors' diagnosis describes a pricing mistake whose bill arrives years later on the vendor's desk rather than NASA's.
Fixed price, small vendors, transferred risk
The deeper criticism targets the fixed-price contract itself. Handing over a service for a set price largely shifts production and cost risk onto vendors that tend to be small and new. The auditors judged that this pressure contributed to one vendor's bankruptcy and continuing market uncertainty for the others, and that as mission scope grew, vendors turned progressively more risk-averse, compounding delays and costs [source: NASA OIG, 2024]. To be fair, the OIG keeps its balance — it still credits CLPS with launching several landers, maintaining NASA oversight while allowing vendor innovation, and stimulating the commercial space economy. In short, CLPS is a model that trades cheapness and speed for risk moved onto the private sector, and the ledger is still open.
Follow that transferred risk one step further and it loops back to NASA. The agency's leverage comes from a pool of 13 eligible companies to choose among — competition is what keeps a delivery fee in the $100 million range [source: NASA, 2026]. Every vendor that fails financially makes that pool smaller. So the audit's finding that one vendor went bankrupt, and that others faced continuing market uncertainty, is not only about those firms [source: NASA OIG, 2024]. It is a note about the durability of the mechanism NASA counts on.
Where the ledger stands
The fair summary is neither vindication nor indictment. Measured against its own plan, CLPS has been slower and more expensive than promised, and hardest on the companies carrying the risk. Measured against what it was set up to do — buy many attempts cheaply instead of one expensively — it has sent several landers toward the Moon at prices in the $100 million range, kept multiple vendors competing, and produced one clean success no U.S. commercial lander had achieved since 1972. Both readings draw on the same evidence; the disagreement is about which yardstick counts.
Conclusion — what to watch
What is settled, and what is not
To recap: in July 2026 NASA committed roughly $600 million to three companies for four lunar deliveries, and all four landers will carry the same navigation, dust, and radiation instruments to weave a data network across the Moon — that much is confirmed by the announcement. Whether those landings actually succeed is a separate question; of four attempts so far, exactly one was a full success. And whether a low-cost commercial model leads to a self-sustaining lunar economy remains a strategic hope and a hypothesis under test.
Three things to watch
So what should you watch? First, whether the late-2028 target holds — given a history of roughly 14-month delays, that date will be tested. Second, whether the identical instrument trio actually works across multiple landing sites and produces the "Moon data network" Kearns described. Third, and biggest, whether private landers can find a market that stands on its own beyond NASA as anchor customer. Whether the commercial Moon economy is taking off or merely cruising will, in the end, be answered by those three.
Underneath all three sits the discipline this article began with: keep the layers apart. The $590 million, the three companies, and the identical instrument trio are settled facts. Late 2028 and the $4.2 billion ceiling are intentions. The one-in-four landing record and the audit's cost and schedule findings are measurements — the only category that tells you how the intentions have fared. The commercial Moon economy is real enough to have a track record. Whether it becomes an economy, rather than a procurement program, is still, in the most literal sense, undelivered.