SpaceX
SpaceX CEO Elon Musk says that BFR could cost less to build than Falcon 9
SpaceX CEO Elon Musk believes that there may be a path for the company to ultimately build the massive Starship spacecraft and Super Heavy booster (formerly BFR) for less than Falcon 9/Falcon Heavy, a rocket 3-9 times smaller than BFR.
While it certainly ranks high on the list of wild and wacky things the CEO has said over the years, there may be a few ways – albeit with healthy qualifications – that Starship/Super Heavy production costs could ultimately compare favorably with SpaceX’s Falcon family of launch vehicles. Nevertheless, there are at least as many ways in which the next-gen rocket can (or should) never be able to beat the production cost of what is effectively a far simpler rocket.
This will sound implausible, but I think there’s a path to build Starship / Super Heavy for less than Falcon 9
— Elon Musk (@elonmusk) February 11, 2019
Dirty boosters done dirt cheap
On the one hand, Musk might not necessarily be wrong, especially if one throws the CEO several bones in the interpretation of his brief tweet. BFR at its simplest is going to require a full 38 main rocket engines to achieve its nominal performance goals, 7 on Starship and 31 on Super Heavy. As a dramatically more advanced, larger, and far more complex engine, Raptor will (with very little doubt) cost far more per engine than the relatively simple Merlin 1D. BFR avionics (flight computers, electronics, wiring, harnesses) are likely to be more of a known quantity, meaning that costs will probably be comparable or even lower than Falcon 9’s when measured as a proportion of overall vehicle cost. Assuming that BFR can use the exact same cold gas thruster assemblies currently flying on Falcon 9, that cost should only grow proportionally with vehicle size. Finally, Starship will not require a deployable payload fairing (~10% of Falcon 9’s production cost).
All of those things mean that Starship/Super Heavy will probably be starting off with far better cost efficiency than Falcon 9 was able to, thanks to almost a decade of interim experience both building, flying, and refurbishing the rocket since its 2010 debut. Still, BFR will have to account for entirely new structures like six large tripod fins/wings and their actuators, wholly new thrust structures (akin to Falcon 9’s octaweb) for both stages, and more. Considering Starship on its own, the production of a human-rated spacecraft capable of safely housing dozens of people in space for weeks or months will almost without a doubt rival the cost of airliner production, where a 737 – with almost half a century of production and flight heritage – still holds a price tag of $100-130+ million.
-
- BFR shown to scale with Falcon 1, 9, and Heavy. (SpaceX)
-
- A September 2018 render of Starship (then BFS) shows one of the vehicle’s two hinged wings/fins/legs. (SpaceX)
-
- BFR’s booster, now known as Super Heavy. (SpaceX)
-
- Sadly, this is a not a sight that will greet Falcon 9 booster B1046’s fourth launch – Crew Dragon’s critical In-Flight Abort test. (SpaceX)
Adding one more assumption, the most lenient interpretation of Musk’s tweet assumes that he is really only subjecting the overall structure (sans engines and any crew-relevant hardware) of BFR relative to Falcon 9. In other words, could a ~300-ton stainless steel rocket structure (BFR) cost the same amount or less to fabricate than a ~30-ton aluminum-lithium alloy rocket structure (Falcon 9/Heavy)? From the very roughest of numerical comparisons, Musk estimated the cost of the stainless steel alloys (300-series) to be used for BFR at around $3 per pound ($6.60/kg), while aluminum-lithium alloys used in aerospace (and on Falcon 9) are sold for around $20/lb ($44/kg)*. As such, simply buying the materials to build the basic structures of BFR and Falcon 9 would cost around and $7.5M and $5M, respectively.
Assuming that the process of assembling, welding, and integrating Starship and Super Heavy structures is somehow 5-10 times cheaper, easier, and less labor-intensive, it’s actually not inconceivable that the cost of building BFR’s structure could ultimately compete with Falcon 9 after production has stabilized after the new rocket’s prototyping phase is over and manufacturing processes are mature.
*Very rough estimate, difficult to find a public cost per unit mass from modern Al-Li suppliers
Costs vs. benefits
On the opposite hand, stainless steel rockets do not have a history of being uniquely cost-effective relative to vehicles using alternative materials. The only orbital-class launch vehicles to use stainless steel (and balloon) tanks are the Atlas booster and the Centaur upper stage, with Atlas dating back to the late 1950s and Centaur beginning launches in the early ’60s. Stainless steel Atlas launches ended in 2005 with the final Atlas III mission, while multiple forms of Centaur continue to fly regularly on ULA’s Atlas V and Delta IV.
Based on a 1966 contract between NASA and General Dynamics placed shortly after Centaur’s tortured development had largely been completed, Centaur upper stages were priced around $25M apiece (2018 USD). In 1980, the hardware for a dedicated Atlas-Centaur launch of a ~1500 kg Comstar I satellite to GTO cost the US the 2018 equivalent of a bit less than $40M ($71M including miscellaneous administrative costs) – $22.4M for Centaur and $17.6M for Atlas. For Atlas, the rocket’s airframe (tanks and general structure) was purchased for around $8.5M. That version of Atlas-Centaur (Atlas-SLV3D Centaur-D1A) was capable of lifting around 5100 kg (11,250 lb) into Low Earth Orbit (LEO) and 1800 kg (~4000 lb) to geostationary transfer orbit (GTO), while it stood around 40m (130 ft) tall, had a tank diameter of 3.05m (10 ft), and weighed ~150t (330,000 lb) fully fueled.
-
- Atlas shows off its shiny steel balloon tanks. (SDASM)
-
- The original space-faring Atlas, known as SM-65, seen here with a Mercury space capsule. (NASA)
-
- A Centaur upper stage is pictured here in 1964. (NASA)
-
- Atlas SLV3D is pictured here launching a Comstar I satellite.
-
- A Falcon 9 booster is seen here near the end of its tank welding, just prior to painting. (SpaceX)
-
- An overview of SpaceX’s Hawthorne factory floor in early 2018. (SpaceX)
In a very loose sense, that particular stainless steel Atlas variant was about half as large and half as capable as the first flight-worthy version of Falcon 9 at roughly the same price at launch ($60-70M). What does this jaunt through the history books tell us about the prospects of a stainless steel Starship and Super Heavy? Well, not much. The problem with trying to understand and pick apart official claims about SpaceX’s next-generation launch architecture is quite simple: only one family of rockets in the history of the industry (Atlas) regularly flew with stainless steel propellant tanks, a half-century lineage that completed its final launch in 2005.
Generally speaking, an industrial sample size of more or less one makes it far from easy to come to any particular conclusions about a given technology or practice, and SpaceX – according to CEO Elon Musk – fully intends to push past the state of the art of stainless steel rocket tankage with BFR. Ultimately, American Marietta/Martin Marietta/Lockheed Martin was never able to produce launch vehicle variants of the stainless steel Atlas family at a cost more than marginally competitive with Falcon 9, despite the latter rocket’s use of a far more expensive metal alloy throughout its primary tanks and structure.
At least 10X cheaper
— Elon Musk (@elonmusk) February 11, 2019
At some point, it’s even worth asking whether the per-unit cost of Starship and Super Heavy should be relevant at all to their design and construction, at least within reason. If the goal of BFR is to drastically lower the cost of launch by radically improving the ease of reuse, it would be truly bizarre (and utterly unintuitive) if those goals could somehow be achieved without dramatically raising the cost of initial hardware procurement. Perhaps the best close comparison to BFR’s goals, modern airliners are eyewateringly expensive ($100-500M apiece) as a consequence of the extraordinary reliability, performance, efficiency, and longevity customers and regulatory agencies demand from them, although those costs are admittedly not the absolute lowest they could be in a perfect manufacturing scenario.
At the end of the day, it appears that Musk is increasingly of the opinion that the pivot to stainless steel could ultimately make BFR simultaneously “better, faster, [&] cheaper”. However improbable that may be, if it does turn out to be the case, Starship and Super Heavy could be an unfathomable leap ahead for reliable and affordable access to space. It could also be another case of Musk’s excitement and optimism getting the better of him and hyping a given product well beyond what it ultimately is able to achieve. Time will tell!
Check out Teslarati’s newsletters for prompt updates, on-the-ground perspectives, and unique glimpses of SpaceX’s rocket launch and recovery processes!
News
SpaceX tells the FCC that Starship Flight 14 is going to orbit
SpaceX filed with the FCC for Starship Flight 14, its first true orbital launch attempt.
SpaceX has asked the Federal Communications Commission for permission to fly Starlink terminals during Starship’s fourteenth flight test, and the filing lays out a genuine trip to orbit, something the program has never attempted.
Every Starship flight so far, including Flight 13’s successful splashdown in the Indian Ocean on July 24, has flown a suborbital arc that ends with the ship reentering the atmosphere within the same hour it launches. The FCC paperwork describes a mission profile built around an actual orbital insertion instead.
The payload is the other half of the story. Flight 13 carried 20 production Starlink V3 satellites, but because that mission never reached orbit, the satellites reentered along with the ship rather than joining the constellation, something Teslarati covered in detail after SpaceX released footage shot from one of those satellites as it drifted away from Starship in space. Flight 14 is designed to close that gap. If the orbital insertion holds, the roughly 20 V3 satellites onboard would separate into an operational orbit and could eventually go into service, each one rated for about 1 terabit per second of downlink capacity by SpaceX’s own account.
SpaceX announces new Starbase for ‘thousands of Starship launches annually’
Elon Musk first flagged the orbital attempt during SpaceX’s August 4 earnings call, the company’s first as a public entity following its June IPO under the ticker SPCX. He also floated catching the ship with the Starbase tower on the same flight, an idea he walked back on August 20, saying the catch attempt would more likely come “in a few months,” as Teslarati reported at the time. Flight 14 will instead target a splashdown for the ship in the Indian Ocean, the same recovery method used since Flight 12.
Hardware has been catching up to the ambition. Booster 21 completed a full 33-engine static fire on August 28, and Ship 41 finished its own six-engine test the week before. An airspace briefing circulated to pilots on August 20 listed September 15 as the target date, later than the end of August window Musk mentioned on the earnings call, though SpaceX has not confirmed a launch date publicly and Starship schedules routinely slip while hardware and FAA paperwork line up.
The FCC filing itself does not guarantee a launch date. It covers communications authority, and not flight readiness, considering SpaceX still needs Ship 41 fully stacked and cleared by the FAA before Flight 14 can fly. But the filing is a real marker of intent and it puts a specific regulatory process behind what had so far only been Musk’s word on the earnings call.
Elon Musk recently restated a fact that still defines SpaceX’s origin story: if Falcon 1’s fourth launch had failed, the company would not exist. The comment answered a reminder that after three consecutive losses, SpaceX had money for only one more attempt.
On X, Peter Diamandis said that the present-day acknowledgement of SpaceX’s success does not discount the rough start the company had. “Almost nobody remembers that Elon’s first rocket failed three times, and there was money for exactly only one more attempt.”
Musk said, “If the 4th launch had failed, SpaceX would not exist.”
If the 4th launch had failed, SpaceX would not exist
— Elon Musk (@elonmusk) August 30, 2026
In late 2008, the firm was nearly out of cash. Another failure would have ended payroll, closed the Hawthorne factory, and left the Falcon 9 and Dragon programs as unfinished drawings.
The first flight lifted off from Omelek Island on 24 March 2006. Thirty-three seconds later, a corroded aluminum fitting on a fuel line leaked. Kerosene ignited around the Merlin engine, control was lost, and the vehicle came apart. The small DARPA payload, FalconSAT-2, survived the short flight only to land on a storage shed near the pad. Investigators later traced the fitting to a materials mix-up that should never have reached the rocket.
Flight 2, on 21 March 2007, looked far better at first. The first stage burned cleanly and handed off to the Kestrel-powered upper stage. The vehicle crossed 100 kilometers and reached a peak of about 289 kilometers. Then propellant slosh in the second-stage tank started a circular coning motion that grew until the engine shut down. Telemetry faded as the stage tumbled, and SpaceX had reached space but not orbit. Over the next year, the team redesigned everything from the ground up, including tanks, baffles, and the new regeneratively cooled Merlin 1C.
That engine flew on Flight 3 on 2 August 2008. The first stage performed almost perfectly and reached 217 kilometers. After main-engine cutoff, leftover fuel in the cooling channels produced a faint residual thrust, roughly 10 pounds per square inch of chamber pressure. On a Texas test stand, the effect was invisible beneath ambient air pressure. In vacuum it was enough to push the spent first stage back into the second stage after separation. The stages collided, the upper stage spun, and the mission was lost. Musk later said a slightly longer delay before staging would have saved the flight.
Six weeks later, the team assembled Flight 4 from remaining parts and flew it on 28 September 2008 at 23:15 UTC. The payload was Ratsat, a 165-kilogram aluminum mass simulator built in-house. Staging was delayed so residual thrust could decay. The Kestrel ignited, the fairing split away, and nine and a half minutes after liftoff the vehicle was in orbit. After a coast, the second stage restarted, settling into a 621-by-643-kilometer path at 9.35 degrees inclination. Falcon 1 became the first privately developed liquid-fueled rocket to reach Earth orbit. Musk called the insertion “middle of the bull’s-eye.”
SpaceX restores a Falcon 1 rocket for 10th anniversary of first launch success
That success unlocked NASA’s Commercial Resupply Services award later that year. Without it, there would have been no Falcon 9, no reusable first stages, and no Dragon cargo or crew flights to the International Space Station. Launch prices would have remained far higher. Starlink’s constellation would not exist; broadband from low Earth orbit would still be a paper concept.
Ride-share markets, high launch cadence, and the current pace of lunar and Mars hardware would be years behind. Communications, Earth observation, and the cost of putting anything into space would look more like the 2000s than the 2020s.
One extra second of residual thrust in August 2008 would have written a different decade.
Elon Musk
OpenAI cites distrust of SpaceX in decision to drop Cursor partnership
OpenAI will cut SpaceX-owned Cursor’s model access in November, citing Musk’s history of broken contracts.
OpenAI, the company behind ChatGPT, announced late Friday that it is ending its partnership with Cursor, cutting off the coding tool’s access to its models on November 12. The move comes two weeks after SpaceX completed its $60 billion acquisition of Cursor’s parent company, Anysphere, folding the popular AI coding assistant into Elon Musk’s growing SpaceXAI division.
In a post on its website, OpenAI said the decision came down to trust, not technology. “We cannot be confident that SpaceX will use our technology within our terms of service, based on our experience with Elon Musk’s companies violating contracts,” the company wrote. OpenAI pointed to two specific incidents: X, now part of SpaceX, allegedly breaking the terms of an existing OpenAI contract after Musk bought Twitter.
That lawsuit is the backdrop for all of this. Musk cofounded OpenAI in 2015, left the board in 2018, and sued Sam Altman and Greg Brockman in 2024, arguing they abandoned the company’s nonprofit mission for profit. A federal jury sided with OpenAI in May, finding Musk waited too long to sue rather than ruling on the merits of his claims. Musk said at the time he would appeal to the Ninth Circuit, calling the outcome a “calendar technicality” rather than a real judgment.
SpaceX’s interest in Cursor predates that verdict by weeks. The company first struck a deal with Cursor in April, securing an option to acquire it for $60 billion or pay $10 billion for joint development work instead. As Teslarati reported at the time, the logic was straightforward: Cursor was paying retail prices to Anthropic and OpenAI, two of its most direct competitors, every time a developer used its product, while SpaceX had idle capacity on its Colossus supercomputer, roughly the equivalent of a million Nvidia H100 GPUs, that Cursor could use to train its own models instead. SpaceX exercised the option in June, days after its own IPO, and the deal closed in mid-August.
Once it closed, Musk moved fast. On an all-hands call with more than 1,000 Cursor employees, he reportedly told staff that SpaceXAI’s Grok was playing catchup in the AI race, unlike Tesla and SpaceX in their own markets, and singled out Anthropic as the company to catch. Cursor CEO Michael Truell now reports directly to Musk inside SpaceXAI.
Losing OpenAI’s models leaves Cursor leaning harder on Anthropic’s Claude, which has its own compute agreement with SpaceX, and on Cursor’s in-house Composer model, the one SpaceX’s compute was supposed to accelerate in the first place. OpenAI framed the November deadline as maximum notice under its contract, and said it wants to “go above and beyond” to help developers through the transition. Whether Anthropic makes the same call is now the open question in AI coding.
SpaceX tells the FCC that Starship Flight 14 is going to orbit
Tesla Cybercab Event: what to expect from Austin
