Starship — the multiplier bet.
Starship is SpaceX's two-stage, fully reusable super-heavy launch vehicle, designed to deliver more than 100 metric tons to low-Earth orbit at marginal cost. It is the single largest capital commitment on the income statement — $3.0 billion of R&D in FY2025 and $930 million in Q1 2026 alone. The day Starship reaches routine commercial service, the cost-per-kilogram delta to every other launch vehicle becomes a step change and Starlink V3 deployment economics reset.
From sub-orbital hops to commercial payload.
Starship's commercial milestones, in order per the S-1:
- Orbital flight test campaign. Multiple integrated flight tests have flown since 2023 with progressively expanded mission profiles — booster catch, ship reentry, on-orbit propellant transfer, controlled splashdowns, payload deployment tests.
- First commercial payload to orbit. Target window: second half of 2026, per the S-1. This is the inflection point the entire valuation case hinges on.
- Starlink V3 deployment at scale. Once Starship is operational, the cost of deploying V2 and V3 satellites drops materially relative to Falcon 9. The S-1 explicitly says V3 satellites cannot be deployed via Falcon at their planned size.
- NASA Artemis HLS. A modified Starship serves as the crewed lunar lander for Artemis III and Artemis IV. This is a contracted revenue line with milestone payments totaling ~$4.0B across both missions.
- Mars cargo and crew. Stated long-term mission. No financial line item attached in the S-1.
Vehicle architecture
| Element | Stat | Notes |
|---|---|---|
| Super Heavy (booster) | 69 m / 33 Raptor engines | Catches on launch tower (Mechazilla) — full reusability |
| Starship (upper) | 50 m / 6 Raptor engines | 3 sea-level + 3 vacuum-optimized |
| Stack height | ~121 m | Tallest, most powerful rocket ever built |
| Payload (reusable) | 100–150 t to LEO | Depends on mission profile |
| Payload (expendable) | ~250 t to LEO | Theoretical max |
| Propellant | LOX / Liquid methane | In-space propellant transfer demonstrated 2025 |
| Production site | Starbase, Boca Chica, TX | Vertical-integrated with Raptor production at McGregor |
Why $3B/year of R&D is the right number.
Starship's economic case rests on three multiplier effects:
1. Cost per kilogram to LEO
Industry benchmarks and SpaceX's own internal targets imply marginal cost-per-kilogram in the $50–200/kg range at full reusability and high cadence, versus Falcon 9's ~$2,000/kg fully-loaded and traditional expendable rockets at $5,000–15,000/kg. The S-1 doesn't commit to a specific number but characterizes the delta as "an order of magnitude" improvement.
2. Starlink V3 deployment
V3 satellites are larger, more capable, and explicitly designed for Starship deployment. Two effects compound:
- Each V3 has more downlink capacity than each V2.
- Each Starship launch can deploy materially more total bandwidth than a Falcon 9 Starlink mission.
The S-1 notes that constellation deployment cost-per-Mbps drops "significantly" with V3 + Starship — the model that underwrites the Starlink margin trajectory at scale.
3. New addressable markets
At 100+ tons to LEO with reusability, several payload classes become economic that aren't today: full-station modules, large telescopes (Apollo-class missions), planetary missions with hosted instrument suites, and point-to-point cargo (a long-term concept, not in S-1 revenue lines).
What can break the Starship story.
- Technical execution. Reaching routine reusable commercial service is non-trivial. The S-1 lists Starship technical execution as the first item in its risk factors. Heat-shield durability across multiple reentries, refurbishment cycles, and ship recovery are the biggest open engineering questions.
- Schedule slip. H2 2026 commercial payload is a target, not a commitment. Past Starship milestones have slipped — orbital flight tests, ship recovery, in-orbit propellant transfer all moved right at various points. A 12-month slip is not a thesis-breaker but compounds the R&D burn.
- Environmental & regulatory. FAA launch licensing, FWS environmental review at Boca Chica, Coast Guard maritime closures, and Texas state permitting are ongoing constraints. A major anomaly or environmental finding could pause cadence.
- R&D burn pre-IPO. The combined Starship + xAI R&D run rate is the reason FY2025 closed with a $4.9B net loss. Until commercial payload service begins, Starship is a cost line, not a revenue line.
- Artemis program changes. NASA's Artemis funding is subject to annual appropriations and political review. A program restructuring could affect HLS milestone payments.
- Raptor production. Starship requires 39 Raptors per stack. Engine production cadence at McGregor is a known constraint that the company has been ramping for years.
For broader risk framing see the risk factors page.