Musk: Starship V4 Grows 40% to ~70m, Adds 3 Vacuum Engines

Elon Musk wrote on X early on Monday that “Starship V4 will be ~70m long, which is roughly a 40% length increase and allows >60% more propellant load.” He added that the “Engine count will be raised to 9, with 3 more Raptor-Vacuum engines added, and these will be higher thrust V4 engines.”

SpaceX’s Starship page lists the ship, the upper stage, at 52 m with six engines and 1,600 t of propellant. Against that listing, ~70m is about 35% longer, and 60% more propellant is 2,560 t, by our arithmetic. Musk’s 40% figure implies a current ship of about 50 m.

Business Pill · Vertical Integration

A short explainer on vertical integration, when a company owns more layers of its own supply chain. It teaches the general idea and does not discuss SpaceX, Starship or this post.

The key insight: As we read SpaceX’s documents, the stretch matters most for SpaceX’s own satellites: its prospectus says Falcon 9 and Falcon Heavy cannot deploy Starlink V3, and Flight 15 carried 26 of the up to 60 V3 satellites the prospectus expects per Starship launch, 43% by our arithmetic.

What Musk Posted

Musk posted at 01:33 UTC on 16 November, quoting a post by @ApoStructura that read “The scale of Starship:”. The post had about 1.16 million views when we read it. It came about 28 hours after SpaceX caught a Starship upper stage with its launch tower for the first time, by our arithmetic.

The post gives three changes: length, propellant and engines. It gives no date, no payload figure and no thrust figure for the new engines beyond “higher thrust V4 engines”.

Starship upper stage as listed on SpaceX’s Starship page (52 m, 1,600 t of propellant, six engines inclu
Starship upper stage as listed on SpaceX’s Starship page (52 m, 1,600 t of propellant, six engines including three Raptor Vacuum) against Musk’s 16 November description of V4. Index values, the 2,560 t floor and the six-vacuum-engine count (which assumes the three sea-level Raptors stay) are our arithmetic.

What Was Posted on X

Elon Musk’s post of 16 November 2026 (01:33 UTC), quoting a post that read “The scale of Starship:”. Musk wrote that Starship V4 will be about 70 m long, roughly 40% longer, with more than 60% more propellant, and that the engine count will rise to nine with three more Raptor-Vacuum engines. This piece sets those figures against SpaceX’s own Starship page.

Musk’s post of 15 November 2026 (22:20 UTC), in which he wrote he had been going over the whole Starship heat shield. He wrote that Ship 42 could probably be reflown with a few weeks of refurbishment, and that many small fixes are needed for airplane-like reusability. The section on the first ship catch quotes it.

The Ship, Not the Stack

The figures fit the upper stage. SpaceX lists the full Starship system at 124 m and the Super Heavy booster at 72 m, “Powered by 33 Raptor engines”. The page says “Starship is powered by six engines, three Raptor engines, and three Raptor Vacuum (RVac) engines, which are designed for use in the vacuum of space.”

Three more Raptor Vacuum engines on a ship with six gives nine, the count Musk gave. If the three sea-level Raptors stay, the vacuum engines double from three to six and the total rises by 50%, by our arithmetic.

The page does not say which version its figures describe. That is why the two baselines differ: 70 m is 35% more than the listed 52 m, while Musk’s roughly 40% points to about 50 m.

What SpaceX Has Said About Payload

SpaceX’s prospectus states that “Starship V3 is designed to deliver 100 metric tons to Earth’s orbit in a fully reusable configuration”. It adds: “Future generations of Starship are being designed to double this payload capacity.” Musk’s post does not tie a payload number to V4.

For scale, SpaceX reported “78 launches and 1,041 metric tons of mass to orbit” in the first half of 2026. That is about 13.3 tons per launch, by our arithmetic. V3’s 100 tons is about 7.5 times that average, and the 200 tons the prospectus describes for future generations is about 15 times.

Starlink Is the First Customer

The prospectus says “Our current operational rockets, including Falcon 9 and Falcon Heavy, are not capable of deploying V3 satellites”. It expects “a single Starship launch will be capable of deploying up to 60 V3 satellites to LEO”, a potential twenty-fold increase in Starlink downlink capacity per launch against Falcon 9.

Flight 15 deployed 26 V3 satellites, which SpaceX says equal about 10 Falcon 9 launches of V2 Mini satellites. That is 43% of the prospectus’s up-to-60 figure, by our arithmetic. Starlink was 81% of Falcon’s 2026 payload mass in our check of Musk’s 85% claim.

On cost, SpaceX’s page says “Starship is designed to deliver payloads farther and at a lower marginal cost per launch than our current Falcon vehicles.” The per-ton price on that page is for cargo to the lunar surface and Mars, “at a rate of $100 million per metric ton”, from 2028 at the earliest.

Rival constellations depend on other rockets. On Sunday a ULA Vulcan failed with 40 Amazon Leo satellites aboard, as we reported.

After the First Ship Catch

SpaceX posted “Mechazilla catches Starship!” at 21:21 UTC on 14 November, the first tower catch of an upper stage, as we reported. The Super Heavy V3 booster on that flight was not caught; our explainer on the tower catches counts four catches in all.

On Sunday evening Musk wrote that “Ship 42 could probably be reflown with a few weeks of refurbishment”, but that “many small fixes are needed to achieve immediate reusability (like an airplane)”.

The Structural Read

As we read the page and the post together, more than 2,560 t of propellant, by our arithmetic, plus three more vacuum engines is a change aimed at the upper stage, the part SpaceX caught for the first time on 14 November.

As we read the prospectus, a 200-ton design goal for future generations would equal about 15 times SpaceX’s first-half 2026 average of 13.3 tons per launch, by our arithmetic; Musk’s post does not say whether V4 is that generation.

Elon Musk on X, 16 November 2026

“Engine count will be raised to 9, with 3 more Raptor-Vacuum engines added, and these will be higher thrust V4 engines.”

Three Implications

FOR STARLINK The prospectus expects up to 60 V3 satellites per Starship launch, a potential twenty-fold increase in downlink capacity deployed against a Falcon 9 launch.

FOR LAUNCH COST SpaceX’s page says Starship is designed to deliver payloads at a lower marginal cost per launch than Falcon; it gives no low-orbit price per ton.

RELATED Our 14 November piece: SpaceX catches Starship from orbit for the first time.

The Business Engineer Lens

This story maps onto the Business Engineer framework The AI Supercycle.

The framework argues: “Each layer is constrained by the one below it.”

As we read it, Starlink V3 is a layer constrained by the rocket below it: SpaceX’s prospectus says its Falcon rockets cannot deploy V3 satellites, so the satellite network’s next step waits on Starship, which SpaceX also builds. That is vertical integration across two layers. The framework predates this post and does not discuss launch.

What Is Not Established

Musk’s post gives no date for V4, no payload and no thrust per engine. SpaceX’s page lists a Raptor at 250 tf but does not label the ship version behind its 52 m, so the 35% and 40% readings cannot be settled from these sources. Whether the three sea-level Raptors stay is our assumption from the engine count.

This Week In The AI Supercycle

AI Lost Its Central Banker, the Lab Oligopoly, and the Future of Open Source

This week’s issue: the Nvidia leadership change, read as the AI build-out losing its central banker: who now finances the GPU landlords, backstops the data centres and sponsors open models, and why two labs are becoming the bank.

Read the full issue →

Business Engineer Framework

The AI Supercycle

The Business Engineer’s map of the layers of AI, from energy and compute to the networks that reach users.

Read the Map of AI →

The Bottom Line

Musk says Starship V4 will be about 70 m long with more than 60% more propellant and nine engines, three more of them Raptor Vacuum. Against SpaceX’s listed 52 m ship with 1,600 t of propellant, that is about 35% longer and more than 2,560 t, by our arithmetic.

95,000+ executives read Business Engineer for the AI strategy frameworks cited by ChatGPT, Claude, and Perplexity.

A note on sourcing. Musk’s posts of 15 and 16 November are quoted from X, which we read in full. Ship and booster dimensions, propellant, engine counts and the cost and price lines are from SpaceX’s Starship vehicle page as read on 16 November; payload and Starlink V3 figures are from SpaceX’s 424B4 prospectus, and launch totals from its second-quarter 2026 results release. The Flight 15 catch and satellite figures are from SpaceX’s posts as reported in our 14 November pieces, the Vulcan failure from our 15 November piece, and the Starlink share from our 16 November analysis of Jonathan McDowell’s catalog. Every percentage, ratio and implied length is our arithmetic. We did not contact SpaceX for this piece.

Sources: Elon Musk on X, 16 Nov 2026 · SpaceX, Starship vehicle page, read 16 Nov 2026 · SpaceX prospectus (Form 424B4), SEC, June 2026 · SpaceX Q2 2026 results release (Ex. 99.1), SEC · Elon Musk on X, 15 Nov 2026

Scroll to Top
FourWeekMBA