SpaceX is trying to fly Starship V3 for the first time today. It's the version that has to work.
After scrubbing earlier this week, SpaceX is targeting today, May 22, for Starship V3's debut (IFT-12): 408 feet, 33 new Raptors, 3× the payload, a new launch pad. It's the redesign that has to make Starship operational.
After scrubbing an attempt earlier this week, SpaceX is targeting today, May 22, for the debut flight of Starship V3 — its 12th integrated flight test, and the first of a clean-sheet redesign that's supposed to turn Starship from a test program into an operational rocket. It lifts off from Starbase in South Texas, on the brand-new Pad 2.
What's actually new in V3
This isn't a tweak; it's close to a new vehicle:
Scale: about 408 feet tall — more powerful than any previous Super Heavy, with 33 redesigned third-generation Raptor engines on the booster and six on the ship.
Payload: more than 100 metric tons to orbit in a fully reusable configuration — roughly three times the previous version.
Propulsion: a "clean-sheet redesign" — a new Raptor startup method, larger propellant tanks, an improved reaction-control system, and a tighter aft section to limit trapped propellant (the failure mode behind earlier flights).
Aero & pad: three larger, stronger grid fins instead of four; and a new launch mount with bigger tanks, faster fueling, and redesigned "chopstick" catch arms.
The flight plan is ambitious for a debut: deploy 22 Starlink satellite simulators, deliberately stress the new flight-control flaps, and attempt the "dynamic banking maneuver" Starship will eventually use to land.
Our read
V3 is the version that has to work.
The earlier Starships proved the architecture could fly and (sometimes) be caught. V3 is meant to prove the economics: three times the payload, fully reusable, flying often. That combination is the entire thesis — Starship only matters if it's cheap and routine, not heroic. It's also the bridge to the business: V3 is what deploys the next generation of Starlink, and Starlink is the cash engine behind the largest IPO in history that SpaceX just filed for. The financial story leans on Starlink today and Starship tomorrow; V3 is where "tomorrow" has to start showing up.
The scrubs are the honest footnote. A first launch of a clean-sheet redesign is statistically as likely to end in a dramatic failure as a clean success, and SpaceX builds that into the plan — its whole method is to fly, break things, and iterate faster than anyone else can. So the number to watch isn't whether this single flight nails every objective. It's the cadence: how quickly V3 flies again, and how fast the fixes land. That turnaround time, more than any one launch, is what tells you whether Starship is becoming a product or staying a prototype.
There's a fitting counterpoint here. The thing actually happening in the sky on May 22 isn't a metaphor or a horoscope — it's 408 feet of stainless steel and 33 engines trying to make orbit cheaper. That's humanity thinking differently about the sky, in the only language that compounds: engineering, flown and reflown until it works.
