> ## Content Index > Fetch the complete content index at: https://www.insidedeeptech.com/llms.txt > Use this file to discover other available public pages before exploring further. # SpaceX Starship V3: A Full Guide to Raptor 3, the First Orbital Flight, Payload Economics, and Honest Limits - URL: https://www.insidedeeptech.com/spacex-starship-v3-full-guide/ - Published: 2026-10-08T06:09:46.000Z - Updated: 2026-10-08T06:09:44.000Z - Description: SpaceX Starship V3 full guide (Oct 2026): specs, Raptor 3, Flights 12 to 14 and the first orbital launch, Starlink V3 payloads, $15 billion of spending, Artemis III, launch sites, and honest limits. - Author: Austin Heaton - Tags: Space, Hardware, AI, Deep Tech On September 28, 2026, a Starship reached orbit for the first time. Flight 14 put Ship 41 into a sustained orbit and released [26 Starlink V3 satellites](https://www.spacex.com/launches/starship-flight-14/?ref=insidedeeptech.com), after SpaceX had invested [over $15 billion in Starship](https://www.sec.gov/Archives/edgar/data/1181412/000162828026040364/spaceexplorationtechnologib.htm?ref=insidedeeptech.com). That flight turned Starship V3 from a test article into a payload-carrying launcher. It also lost an engine on each stage and came home after about three hours instead of the planned ten ([Reuters via The Star](https://www.thestar.com.my/news/world/2026/09/28/spacex039s-starship-launches-on-14th-flight-first-headed-to-orbit?ref=insidedeeptech.com)). This full guide covers what changed in Starship V3, how Raptor 3 works, what Flights 12, 13 and 14 actually proved, the Starlink V3 payload it was built around, the money, Artemis III, launch sites, and the honest limits. For the wider market around it, see Inside Deep Tech's [7 space tech trends reshaping business in 2026](https://www.insidedeeptech.com/7-space-tech-trends-reshaping-business-in-2026/). Last updated: October 8, 2026. ## Key Takeaways - Starship V3 stands 124 m tall and targets [100+ tonnes to orbit, fully reusable](https://www.spacex.com/vehicles/starship/technical-specifications?ref=insidedeeptech.com). - Flight 14 reached orbit and deployed [26 Starlink V3 satellites on September 28](https://www.spacex.com/launches/starship-flight-14/?ref=insidedeeptech.com). - Each Starlink V3 is designed for [1 Tbps of downlink, about 10x a V2](https://convergedigest.com/spacex-starship-deploys-26-starlink-v3-satellites/?ref=insidedeeptech.com). - SpaceX has put [over $15 billion into Starship](https://www.sec.gov/Archives/edgar/data/1181412/000162828026040364/spaceexplorationtechnologib.htm?ref=insidedeeptech.com), with ship catch and refueling still ahead. - Artemis III is now a [late-2027 Earth-orbit docking test](https://arstechnica.com/space/2026/04/put-it-in-pencil-nasas-artemis-iii-mission-will-launch-no-earlier-than-late-2027/?ref=insidedeeptech.com), not a Moon landing. ⚠️ Two different things are called V3\. Starship V3 (also called Block 3) is the rocket. Starlink V3 is the broadband satellite it carries. Flight 14 was the first time both flew together to orbit. ## What Starship V3 is Starship is a two-stage, fully reusable system: the Super Heavy booster and the Starship upper stage, which is also the spacecraft. V3 is the third major version, and it's the configuration SpaceX's IPO filing describes as designed to deliver [100 metric tons to orbit while fully reusable](https://www.sec.gov/Archives/edgar/data/1181412/000162828026040364/spaceexplorationtechnologib.htm?ref=insidedeeptech.com). V3 debuted on Flight 12 on May 22, 2026, which was also the first flight of Raptor 3 and the first launch from Starbase's second pad ([SpaceX](https://www.spacex.com/launches/starship-flight-12?ref=insidedeeptech.com)). By SpaceX's own count, that was the program's twelfth flight test since 2023. The headline change is margin. V2 vehicles demonstrated closer to [35 tonnes of payload](https://en.wikipedia.org/wiki/SpaceX%5FStarship?ref=insidedeeptech.com) on suborbital profiles. V3 stretches the tanks, swaps in a lighter engine and adds the hardware needed for docking and refueling. ### Starship V3 specifications at a glance | Parameter | Starship V3 value | Source | | ---------------------- | ----------------------------------------------- | -------------------------------------------------------------------------------------------------- | | Full stack height | 124 m (407 ft) | [SpaceX](https://www.spacex.com/vehicles/starship/technical-specifications?ref=insidedeeptech.com) | | Diameter | 9 m (29.5 ft) | [SpaceX](https://www.spacex.com/vehicles/starship/technical-specifications?ref=insidedeeptech.com) | | Payload to orbit | 100+ t, fully reusable | [SpaceX](https://www.spacex.com/vehicles/starship/technical-specifications?ref=insidedeeptech.com) | | Super Heavy height | 72 m (236 ft) | [SpaceX](https://www.spacex.com/vehicles/starship/technical-specifications?ref=insidedeeptech.com) | | Super Heavy propellant | 3,650 t (8 Mlb) | [SpaceX](https://www.spacex.com/vehicles/starship/technical-specifications?ref=insidedeeptech.com) | | Super Heavy thrust | 8,240 tf (18.1 Mlbf), 33 Raptor engines | [SpaceX](https://www.spacex.com/vehicles/starship/technical-specifications?ref=insidedeeptech.com) | | Ship height | 52 m (171 ft) | [SpaceX](https://www.spacex.com/vehicles/starship/technical-specifications?ref=insidedeeptech.com) | | Ship propellant | 1,600 t (3.5 Mlb) | [SpaceX](https://www.spacex.com/vehicles/starship/technical-specifications?ref=insidedeeptech.com) | | Ship thrust | 1,614 tf (3.5 Mlbf), 3 Raptor + 3 Raptor Vacuum | [SpaceX](https://www.spacex.com/vehicles/starship/technical-specifications?ref=insidedeeptech.com) | | Raptor thrust (listed) | 250 tf (551 klbf) per engine | [SpaceX](https://www.spacex.com/vehicles/starship/technical-specifications?ref=insidedeeptech.com) | Source: [SpaceX Starship technical specifications](https://www.spacex.com/vehicles/starship/technical-specifications?ref=insidedeeptech.com), as published in October 2026. ## Raptor 3: the engine that makes V3 possible Raptor is a methane-oxygen, full-flow staged-combustion engine. Super Heavy carries 33 of them, with 13 gimbaling engines in the center and 20 around the edge, and SpaceX says each one has roughly [twice the thrust of a Falcon 9 Merlin](https://www.spacex.com/vehicles/starship/technical-specifications?ref=insidedeeptech.com). Raptor 3's biggest gain is mass, not thrust. The S-1 says each Raptor 3 saves [nearly a ton of vehicle mass](https://www.sec.gov/Archives/edgar/data/1181412/000162828026040364/spaceexplorationtechnologib.htm?ref=insidedeeptech.com) versus earlier versions, by removing engine heat shields and simplifying plumbing. Multiply that across 39 engines on a full stack and the payload math changes. Be careful with thrust figures. Some outlets cite about [280 tonnes of thrust per Raptor 3](https://newspaceeconomy.ca/2026/04/16/detailed-review-of-starship-v3/?ref=insidedeeptech.com), but SpaceX's own specification page lists [250 tf per engine and 8,240 tf for the booster](https://www.spacex.com/vehicles/starship/technical-specifications?ref=insidedeeptech.com). Inside Deep Tech uses SpaceX's published numbers until it updates them. Reliability is the open question. On Flights 12 and 14, one Raptor shut down early on each stage during ascent ([Flight 12](https://www.spacex.com/launches/starship-flight-12?ref=insidedeeptech.com), [Flight 14](https://www.spacex.com/launches/starship-flight-14/?ref=insidedeeptech.com)). Engine-out tolerance worked each time, but a booster that's meant to fly like an airliner can't keep losing engines. ## What Flights 12, 13 and 14 actually proved | Flight | Date | Payload | Key result | Problems | Source | | --------- | ------------- | --------------------------- | -------------------------------------------------------------------- | ----------------------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ | | Flight 12 | May 22, 2026 | Starlink mass simulators | First V3 and Raptor 3 flight; ship splashed down in the Indian Ocean | Booster failed boostback relight, hard splashdown; FAA mishap | [SpaceX](https://www.spacex.com/launches/starship-flight-12?ref=insidedeeptech.com), [Wikipedia](https://en.wikipedia.org/wiki/Starship%5Fflight%5Ftest%5F12?ref=insidedeeptech.com) | | Flight 13 | July 24, 2026 | 20 Starlink V3 (suborbital) | All 20 deployed and contacted; in-space relight; intact splashdown | Earlier July 16 attempt aborted when 4 Raptors failed to ignite | [SpaceX](https://www.spacex.com/launches/starship-flight-13?ref=insidedeeptech.com), [Astronomy](https://www.astronomy.com/space-exploration/spacex-launches-starship-test-flight-lucky-number-13/?ref=insidedeeptech.com) | | Flight 14 | Sept 28, 2026 | 26 Starlink V3 (orbital) | First orbit; first deorbit burn; on-target Pacific splashdown | 1 booster engine and 1 Raptor Vacuum shut down; mission cut short | [SpaceX](https://www.spacex.com/launches/starship-flight-14/?ref=insidedeeptech.com), [Reuters](https://www.thestar.com.my/news/world/2026/09/28/spacex039s-starship-launches-on-14th-flight-first-headed-to-orbit?ref=insidedeeptech.com) | Source: [SpaceX Flight 12](https://www.spacex.com/launches/starship-flight-12?ref=insidedeeptech.com), [Flight 13](https://www.spacex.com/launches/starship-flight-13?ref=insidedeeptech.com) and [Flight 14](https://www.spacex.com/launches/starship-flight-14/?ref=insidedeeptech.com) mission pages; [Reuters via The Star](https://www.thestar.com.my/news/world/2026/09/28/spacex039s-starship-launches-on-14th-flight-first-headed-to-orbit?ref=insidedeeptech.com); [Astronomy](https://www.astronomy.com/space-exploration/spacex-launches-starship-test-flight-lucky-number-13/?ref=insidedeeptech.com). ### Flight 12: a good ship and a lost booster The ship did its job. It lost one Raptor Vacuum on ascent, still hit its planned trajectory, stressed its rear flaps on purpose during reentry, and landed on two engines in the Indian Ocean ([SpaceX](https://www.spacex.com/launches/starship-flight-12?ref=insidedeeptech.com)). The booster didn't. Super Heavy failed to relight properly for its boostback burn and hit the Gulf hard ([Wikipedia](https://en.wikipedia.org/wiki/Starship%5Fflight%5Ftest%5F12?ref=insidedeeptech.com)), and the FAA classified the flight as a mishap. The S-1 confirms several Starship flights, [including the 12th](https://www.sec.gov/Archives/edgar/data/1181412/000162828026040364/spaceexplorationtechnologib.htm?ref=insidedeeptech.com), drew FAA mishap determinations, with return to flight conditioned on corrective actions. ### Flight 13: the first real Starlink V3s Flight 13 deployed [20 Starlink V3 satellites](https://www.spacex.com/launches/starship-flight-13?ref=insidedeeptech.com), and SpaceX made contact with each one by radio and laser link before they reentered about 20 minutes later, as planned. The ship relit an engine in space and came down intact in the Indian Ocean ([Astronomy](https://www.astronomy.com/space-exploration/spacex-launches-starship-test-flight-lucky-number-13/?ref=insidedeeptech.com)). After the flight, Elon Musk said SpaceX would try to catch the ship with the tower on the next flight if the data held up ([Astronomy](https://www.astronomy.com/space-exploration/spacex-launches-starship-test-flight-lucky-number-13/?ref=insidedeeptech.com)). Flight 14 flew an orbital mission and splashed down instead, so the ship catch is still pending. ### Flight 14: orbit, with an asterisk Flight 14 lifted off at 7:48 a.m. CT on September 28\. A single sea-level Raptor performed the orbit insertion burn, and Starship deployed all [26 Starlink V3 satellites](https://www.spacex.com/launches/starship-flight-14/?ref=insidedeeptech.com), roughly 269 km up, according to [Reuters](https://www.thestar.com.my/news/world/2026/09/28/spacex039s-starship-launches-on-14th-flight-first-headed-to-orbit?ref=insidedeeptech.com). The asterisk is duration. The plan called for about six orbits and a splashdown after roughly 9 hours 50 minutes, but after the engine issue SpaceX ended the flight after about [3 hours 8 minutes](https://marsprogress.com/starship/flight-14/?ref=insidedeeptech.com). It still flew Starship's first deorbit burn and splashed down on target in the North Pacific ([SpaceX](https://www.spacex.com/launches/starship-flight-14/?ref=insidedeeptech.com)). ## Starlink V3: the payload V3 was built to carry Starship V3 and Starlink V3 are one business case. The S-1 says SpaceX's Falcon 9 and Falcon Heavy [can't deploy V3 satellites](https://www.sec.gov/Archives/edgar/data/1181412/000162828026040364/spaceexplorationtechnologib.htm?ref=insidedeeptech.com), so the next generation of Starlink capacity depends entirely on Starship flying often. | Starlink V3 spec | Value | Source | | ---------------------- | --------------------------------------------------- | ---------------------------------------------------------------------------------------------------------------------------------------------------- | | Downlink per satellite | 1 Tbps (about 10x V2) | [Converge Digest](https://convergedigest.com/spacex-starship-deploys-26-starlink-v3-satellites/?ref=insidedeeptech.com) | | Uplink per satellite | 160 Gbps (about 22x V2) | [Converge Digest](https://convergedigest.com/spacex-starship-deploys-26-starlink-v3-satellites/?ref=insidedeeptech.com) | | User beams | 2,048 downlink and 2,048 uplink | [Converge Digest](https://convergedigest.com/spacex-starship-deploys-26-starlink-v3-satellites/?ref=insidedeeptech.com) | | Laser links | 6 x 400 Gbps optical inter-satellite links | [Converge Digest](https://convergedigest.com/spacex-starship-deploys-26-starlink-v3-satellites/?ref=insidedeeptech.com) | | Planned batch size | About 60 per Starship launch | [Reuters](https://www.thestar.com.my/news/world/2026/09/28/spacex039s-starship-launches-on-14th-flight-first-headed-to-orbit?ref=insidedeeptech.com) | | Capacity per launch | More than 20x a Falcon 9 V2 Mini launch, per SpaceX | [Converge Digest](https://convergedigest.com/spacex-starship-deploys-26-starlink-v3-satellites/?ref=insidedeeptech.com) | Source: [Converge Digest](https://convergedigest.com/spacex-starship-deploys-26-starlink-v3-satellites/?ref=insidedeeptech.com), summarizing SpaceX's published Starlink V3 figures, and [Reuters via The Star](https://www.thestar.com.my/news/world/2026/09/28/spacex039s-starship-launches-on-14th-flight-first-headed-to-orbit?ref=insidedeeptech.com). By Inside Deep Tech's arithmetic, Flight 14's 26 satellites carried about [26 Tbps of designed downlink capacity](https://convergedigest.com/spacex-starship-deploys-26-starlink-v3-satellites/?ref=insidedeeptech.com). That's design capacity, not live throughput: each satellite still has to raise its orbit and pass checkout before it serves customers ([SpaceX](https://www.spacex.com/launches/starship-flight-14/?ref=insidedeeptech.com)). The base it lands on is already large. As of March 31, 2026, SpaceX operated about [9,600 Starlink satellites serving 10.3 million subscribers](https://www.sec.gov/Archives/edgar/data/1181412/000162828026040364/spaceexplorationtechnologib.htm?ref=insidedeeptech.com) in 164 countries, territories and other markets. ## How Starship V3 compares with Falcon and V2 | Vehicle | Payload to LEO | Reuse | Status, Oct 2026 | Source | | --------------------------- | ----------------------- | ----------------------------------- | ------------------------------ | ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | Falcon 9 | About 23 t (expendable) | Booster reused up to 34 times | About 620 launches by Mar 2026 | [SpaceX S-1](https://www.sec.gov/Archives/edgar/data/1181412/000162828026040364/spaceexplorationtechnologib.htm?ref=insidedeeptech.com) | | Falcon Heavy | About 64 t | Partially reusable | Operational since 2018 | [SpaceX S-1](https://www.sec.gov/Archives/edgar/data/1181412/000162828026040364/spaceexplorationtechnologib.htm?ref=insidedeeptech.com) | | Starship V2 (Block 2) | About 35 t | Designed reusable | Retired, never reached orbit | [Wikipedia](https://en.wikipedia.org/wiki/SpaceX%5FStarship?ref=insidedeeptech.com) | | Starship V3 (Block 3) | 100+ t, fully reusable | Booster and ship designed for catch | Reached orbit on Flight 14 | [SpaceX](https://www.spacex.com/vehicles/starship/technical-specifications?ref=insidedeeptech.com), [SpaceX](https://www.spacex.com/launches/starship-flight-14/?ref=insidedeeptech.com) | | Future Starship generations | About 2x V3 | Fully reusable | In design | [SpaceX S-1](https://www.sec.gov/Archives/edgar/data/1181412/000162828026040364/spaceexplorationtechnologib.htm?ref=insidedeeptech.com) | Source: [SpaceX S-1/A](https://www.sec.gov/Archives/edgar/data/1181412/000162828026040364/spaceexplorationtechnologib.htm?ref=insidedeeptech.com), [SpaceX Starship specifications](https://www.spacex.com/vehicles/starship/technical-specifications?ref=insidedeeptech.com), [SpaceX Flight 14](https://www.spacex.com/launches/starship-flight-14/?ref=insidedeeptech.com), [Wikipedia](https://en.wikipedia.org/wiki/SpaceX%5FStarship?ref=insidedeeptech.com). Here's why that matters. V3's reusable target is more than [4x Falcon 9's expendable payload](https://www.sec.gov/Archives/edgar/data/1181412/000162828026040364/spaceexplorationtechnologib.htm?ref=insidedeeptech.com), and it's meant to bring both stages back. If that works at cadence, it's a different cost curve, not a bigger rocket on the same one. ## The money behind Starship V3 SpaceX's IPO filing gives the clearest view yet of the spend. Space-segment R&D rose to [$3,004 million in 2025 from $1,835 million in 2024](https://www.sec.gov/Archives/edgar/data/1181412/000162828026040364/spaceexplorationtechnologib.htm?ref=insidedeeptech.com), which the company attributes to accelerating Starship development, on top of more than $15 billion invested in the vehicle overall. The goal is a cost collapse. The S-1 says Starship aims to cut the cost of reaching orbit by [99% or more](https://www.sec.gov/Archives/edgar/data/1181412/000162828026040364/spaceexplorationtechnologib.htm?ref=insidedeeptech.com) against a historical average of $18,500 per kilogram, a figure it cites from NASA. The same filing says Falcon 9 already brought that to about $2,700 per kilogram in 2010. SpaceX has also started pricing deep-space cargo. Lunar and Mars surface cargo flights for research start no earlier than 2028 at [$100 million per metric ton](https://www.spacex.com/vehicles/starship/technical-specifications?ref=insidedeeptech.com). That's a list price for a service that hasn't flown, so treat it as an anchor, not a quote. 💡 Inside Deep Tech's take: Flight 14 proves Starship V3 can get to orbit with a real payload. It doesn't prove the economics. The S-1's 99% cost-cut claim only holds with both stages caught and reflown fast, and as of October 2026 V3 has caught neither stage. Price your plans on demonstrated reuse, not on the target. For how investors are pricing the rest of the sector around SpaceX, read the team's [guide to current space tech investment trends](https://www.insidedeeptech.com/a-guide-to-current-space-tech-investment-trends/). [See the space tech investment trends](https://www.insidedeeptech.com/a-guide-to-current-space-tech-investment-trends/) ## Starship V3 and Artemis: what changed in 2026 Starship is NASA's Human Landing System for Artemis, which the S-1 describes as built to [deliver astronauts and cargo to the lunar surface](https://www.sec.gov/Archives/edgar/data/1181412/000162828026040364/spaceexplorationtechnologib.htm?ref=insidedeeptech.com). The mission plan around it moved a lot this year. In February 2026, NASA dropped the south-pole landing from Artemis III. In April, Administrator Jared Isaacman told lawmakers both SpaceX and Blue Origin could support a [late-2027 rendezvous and docking test in Earth orbit](https://arstechnica.com/space/2026/04/put-it-in-pencil-nasas-artemis-iii-mission-will-launch-no-earlier-than-late-2027/?ref=insidedeeptech.com), ahead of a landing attempt in 2028\. Track the official plan on NASA's [Artemis III mission page](https://www.nasa.gov/mission/artemis-iii/?ref=insidedeeptech.com). That reshuffle helps SpaceX. Both landers must be refueled in orbit to reach the Moon, but an Earth-orbit mission [doesn't need refueling](https://arstechnica.com/space/2026/04/put-it-in-pencil-nasas-artemis-iii-mission-will-launch-no-earlier-than-late-2027/?ref=insidedeeptech.com). It buys time for the hardest part of the architecture. Refueling is still unproven. The S-1 lists cryogenic propellant transfer in microgravity, propellant settling and boil-off as [technical complexities required for lunar and interplanetary missions](https://www.sec.gov/Archives/edgar/data/1181412/000162828026040364/spaceexplorationtechnologib.htm?ref=insidedeeptech.com). No ship-to-ship transfer had flown as of October 8, 2026. ## Launch sites, licenses and cadence Cadence is the real product. At Starbase, Texas, the FAA's licensing action covers up to [25 Starship launches a year](https://www.faa.gov/space/stakeholder%5Fengagement/spacex%5Fstarship?ref=insidedeeptech.com), with up to 25 landings for each stage. Florida is the expansion path. The FAA's environmental review for Kennedy Space Center's LC-39A analyzed up to [44 Starship launches a year](https://www.faa.gov/space/stakeholder%5Fengagement/spacex%5Fstarship%5Fksc?ref=insidedeeptech.com), with up to 44 landings each for booster and ship. No V3 has launched from Florida yet. Production is being sized to match. SpaceX says Starfactory is designed to build up to [1,000 Starships per year](https://www.spacex.com/vehicles/starship/technical-specifications?ref=insidedeeptech.com), and the S-1 describes a Gigabay with [24 work cells and 400-ton cranes](https://www.sec.gov/Archives/edgar/data/1181412/000162828026040364/spaceexplorationtechnologib.htm?ref=insidedeeptech.com) for integration and refurbishment. ## Why AI investors watch Starship V3 The S-1 ties Starship directly to AI. SpaceX says it expects to begin deploying orbital AI compute satellites, solar-powered and space-cooled, [as early as 2028](https://www.sec.gov/Archives/edgar/data/1181412/000162828026040364/spaceexplorationtechnologib.htm?ref=insidedeeptech.com), and the same filing says Starship's capacity is what enables orbital AI compute. That's a big claim with no flight hardware yet. For the physics, the cooling problem and the economics of orbital compute, read Inside Deep Tech's analysis of data centers in space. [Data Centers in Space: Hype or the Next Infrastructure Frontier?Space data centers leverage solar energy and vacuum cooling to scale compute capacity. Inside Deep Tech looks at why orbital infrastructure matters.![](https://www.insidedeeptech.com/favicon.ico)Inside Deep Tech](https://www.insidedeeptech.com/data-centers-in-space-hype-or-the-next-infrastructure-frontier/) ## What comes next for Starship V3 SpaceX's own filing names the next two milestones: [catching the upper stage and demonstrating in-orbit propellant transfer](https://www.sec.gov/Archives/edgar/data/1181412/000162828026040364/spaceexplorationtechnologib.htm?ref=insidedeeptech.com). Both are needed before V3 is the rapidly reusable rocket the business plan assumes. - **Ship catch.** Musk said a tower catch of the ship was the [next goal after Flight 13](https://www.astronomy.com/space-exploration/spacex-launches-starship-test-flight-lucky-number-13/?ref=insidedeeptech.com). It hasn't happened yet. - **Booster return.** No V3 booster has returned to the tower; Flight 12's [hard splashdown](https://www.spacex.com/launches/starship-flight-12?ref=insidedeeptech.com) set the tone, and Flight 14's booster splashed in the [Gulf](https://www.spacex.com/launches/starship-flight-14/?ref=insidedeeptech.com). - **Propellant transfer.** A ship-to-ship demo is the gate for [lunar Starship missions](https://arstechnica.com/space/2026/04/put-it-in-pencil-nasas-artemis-iii-mission-will-launch-no-earlier-than-late-2027/?ref=insidedeeptech.com). - **Florida debut.** The first LC-39A launch would test the [44-launch-a-year](https://www.faa.gov/space/stakeholder%5Fengagement/spacex%5Fstarship%5Fksc?ref=insidedeeptech.com) East Coast plan. - **Bigger batches.** Moving from 26 to about [60 Starlink V3s per flight](https://www.thestar.com.my/news/world/2026/09/28/spacex039s-starship-launches-on-14th-flight-first-headed-to-orbit?ref=insidedeeptech.com) is the commercial proof point. ## Honest limits of Starship V3 The real limits as of October 2026: - **Engine reliability.** Flights 12 and 14 each lost a Raptor on both stages during ascent ([Flight 12](https://www.spacex.com/launches/starship-flight-12?ref=insidedeeptech.com), [Flight 14](https://www.spacex.com/launches/starship-flight-14/?ref=insidedeeptech.com)), and a [July 16 Flight 13 attempt aborted](https://www.astronomy.com/space-exploration/spacex-launches-starship-test-flight-lucky-number-13/?ref=insidedeeptech.com) when four Raptors failed to ignite. - **No reuse yet.** Neither V3 stage has been caught or reflown. The [99% cost-cut goal](https://www.sec.gov/Archives/edgar/data/1181412/000162828026040364/spaceexplorationtechnologib.htm?ref=insidedeeptech.com) depends on both. - **One orbital flight.** Flight 14 was cut from about ten hours to [about three](https://www.thestar.com.my/news/world/2026/09/28/spacex039s-starship-launches-on-14th-flight-first-headed-to-orbit?ref=insidedeeptech.com). One success is not a track record. - **Refueling is unflown.** Lunar landers [need in-orbit refueling](https://arstechnica.com/space/2026/04/put-it-in-pencil-nasas-artemis-iii-mission-will-launch-no-earlier-than-late-2027/?ref=insidedeeptech.com), and the S-1 flags it as a [key technical risk](https://www.sec.gov/Archives/edgar/data/1181412/000162828026040364/spaceexplorationtechnologib.htm?ref=insidedeeptech.com). - **Regulatory pace.** Each mishap brings an [FAA investigation](https://www.sec.gov/Archives/edgar/data/1181412/000162828026040364/spaceexplorationtechnologib.htm?ref=insidedeeptech.com), and cadence is capped by licenses of [25 a year at Starbase](https://www.faa.gov/space/stakeholder%5Fengagement/spacex%5Fstarship?ref=insidedeeptech.com) and [44 at LC-39A](https://www.faa.gov/space/stakeholder%5Fengagement/spacex%5Fstarship%5Fksc?ref=insidedeeptech.com). - **Self-reported data.** Most performance numbers come from SpaceX's [own pages](https://www.spacex.com/vehicles/starship/technical-specifications?ref=insidedeeptech.com) and [filing](https://www.sec.gov/Archives/edgar/data/1181412/000162828026040364/spaceexplorationtechnologib.htm?ref=insidedeeptech.com). Independent figures, like [280 tf Raptor 3 thrust](https://newspaceeconomy.ca/2026/04/16/detailed-review-of-starship-v3/?ref=insidedeeptech.com), don't always match. 🚀 Inside Deep Tech's take: Starship V3 is now the most important launch vehicle in the world, because the Starlink, Artemis and orbital-compute plans all route through it. Expect the next 12 months to be decided by two boring-sounding events, a ship catch and a propellant transfer. Until both fly, plan around V3 as a powerful expendable-ish rocket with reuse upside, not as an airliner. --- ## Frequently asked questions #### What is new in Starship V3? Starship V3 adds Raptor 3 engines, stretched tanks and a 124 m stack built for 100+ tonnes to orbit fully reusable ([SpaceX](https://www.spacex.com/vehicles/starship/technical-specifications?ref=insidedeeptech.com)). Each Raptor 3 saves nearly a ton of mass versus earlier engines ([SpaceX S-1](https://www.sec.gov/Archives/edgar/data/1181412/000162828026040364/spaceexplorationtechnologib.htm?ref=insidedeeptech.com)). #### Has Starship V3 reached orbit? Yes. Flight 14 reached orbit on September 28, 2026 and deployed 26 Starlink V3 satellites ([SpaceX](https://www.spacex.com/launches/starship-flight-14/?ref=insidedeeptech.com)). The mission was shortened after an engine shut down early. #### How much can Starship V3 carry? SpaceX designs Starship V3 to carry more than 100 metric tons to orbit in a fully reusable configuration ([SpaceX](https://www.spacex.com/vehicles/starship/technical-specifications?ref=insidedeeptech.com)). Future generations are being designed to roughly double that ([SpaceX S-1](https://www.sec.gov/Archives/edgar/data/1181412/000162828026040364/spaceexplorationtechnologib.htm?ref=insidedeeptech.com)). #### How many engines does Starship V3 have? Starship V3 uses 39 Raptor engines in total. Super Heavy has 33, and the ship has 3 sea-level Raptors plus 3 Raptor Vacuum engines ([SpaceX](https://www.spacex.com/vehicles/starship/technical-specifications?ref=insidedeeptech.com)). #### How much has SpaceX spent on Starship? SpaceX says it has invested over $15 billion in Starship ([SpaceX S-1](https://www.sec.gov/Archives/edgar/data/1181412/000162828026040364/spaceexplorationtechnologib.htm?ref=insidedeeptech.com)). Space-segment R&D reached $3,004 million in 2025, driven by faster Starship development. #### Will Starship V3 land astronauts on the Moon? Not on Artemis III. That mission is now a late-2027 docking test in Earth orbit, with a landing attempt planned for 2028 ([Ars Technica](https://arstechnica.com/space/2026/04/put-it-in-pencil-nasas-artemis-iii-mission-will-launch-no-earlier-than-late-2027/?ref=insidedeeptech.com)). #### When will Starship catch the ship with the tower? SpaceX hasn't set a date. It lists catching the upper stage as a key upcoming milestone in its [IPO filing](https://www.sec.gov/Archives/edgar/data/1181412/000162828026040364/spaceexplorationtechnologib.htm?ref=insidedeeptech.com), and Musk had targeted the flight after Flight 13 ([Astronomy](https://www.astronomy.com/space-exploration/spacex-launches-starship-test-flight-lucky-number-13/?ref=insidedeeptech.com)). ## What to watch next Three signals will tell you if Starship V3 is ready to carry real business plans: a tower catch of the ship, a ship-to-ship propellant transfer, and a full batch of about [60 Starlink V3 satellites on one flight](https://www.thestar.com.my/news/world/2026/09/28/spacex039s-starship-launches-on-14th-flight-first-headed-to-orbit?ref=insidedeeptech.com). Each one is on SpaceX's own list in the [S-1](https://www.sec.gov/Archives/edgar/data/1181412/000162828026040364/spaceexplorationtechnologib.htm?ref=insidedeeptech.com). If your company depends on launch, from satellite operators to orbital compute startups, model two cases this quarter: Starship flying monthly with reuse, and Starship flying a few times a year without it. Then see who's already building for the first case. [See the space tech startups to watch](https://www.insidedeeptech.com/best-space-tech-startups-to-watch-in-2026/)