Starlink Satellite Internet for Aviation [2026 Price & Status]

Starlink Aviation is SpaceX’s in-flight connectivity service: a flat-panel, electronically steered antenna bolted to the aircraft crown that talks to low-Earth-orbit satellites at roughly 550km instead of geostationary birds at 35,786km. That single design choice is why it delivers 20-40ms latency and gate-to-gate coverage rather than the 600ms-plus, cruise-only experience most passengers still associate with plane WiFi. As of September 2026, about 48 airlines worldwide have committed to the system — 23 of them signing during 2026 — and the newest installations are rated above 2Gbps of total bandwidth per aircraft.

It is worth being clear about what this product actually is, because the aviation branding obscures it: this is the same satellite internet service that covers homes, ships, RVs and remote worksites, running over the same constellation, with a hardened terminal and an airworthiness certificate wrapped around it. The satellites overhead a Gulfstream at FL450 are the satellites overhead a rural dish. What changes is the terminal, the certification burden and the price — and those three variables explain almost every question people ask about Starlink availability and Starlink price in the air.

The short answer to the question most readers actually have: whether Starlink is on your flight in 2026 depends less on the airline than on the individual tail number. Fleet retrofits run for years, and most carriers are mixing Starlink-equipped aircraft with legacy Gogo and Viasat hardware on the same routes, often on the same day.

What Is Starlink Aviation and How Does It Work

Starlink Aviation is a managed connectivity service sold by SpaceX to aircraft operators, bundling the Aero terminal hardware, the satellite capacity and the network operations behind it. The airborne hardware is the Aero terminal: a low-profile flat panel measuring approximately 75cm x 50cm, mounted under a radome on the fuselage crown. It carries no moving parts in the aperture. Instead of mechanically pointing a dish, it uses an electronically steered phased array — thousands of radiating elements whose signals are phase-shifted in software to form and sweep a beam in microseconds.

That distinction is the engineering heart of the product. A mechanically steered GEO antenna has to physically slew to track a satellite as the aircraft banks, climbs and turns, and it only ever has one target to look at. A phased array can hand off between fast-moving LEO satellites without mechanical lag, which is what makes coverage possible on the taxiway, through the climb and on descent, not just in the cruise.

LEO versus GEO in one paragraph

Geostationary satellites sit at 35,786km above the equator. Radio waves take roughly 120 milliseconds to travel that distance one way, so a round trip — request out, response back — cannot physically beat about half a second, before any processing. Starlink’s operational shell sits near 550km, a factor of 65 closer. The physics cascade from there: lower round-trip delay, higher usable throughput per unit of spectrum, smaller and cheaper user terminals, and a beam footprint small enough that capacity is not shared across an entire continent.

• Orbit: Starlink at approximately 550km; Gogo 2Ku and Viasat at 35,786km.
• Antenna: electronically steered phased array, roughly 75cm x 50cm, no moving aperture.
• Coverage profile: gate-to-gate, enabled by LEO satellite handoff at all altitudes — including on the ground.
• Backhaul: inter-satellite laser links on v1.5 and later satellites move traffic between spacecraft, removing the need for a ground station within the satellite’s view.
• Operating envelope: SpaceX publishes a 480 km/h ceiling for its land-mobility terminals, which does not apply to aviation hardware. Aviation terminals are cleared to the flight envelope of the specific airframe through the Supplemental Type Certificate covering that installation, not through a single published speed figure.

In practice, the passenger-visible consequence of gate-to-gate capability is mundane and enormous: the network is live while you are still boarding, and it does not drop at 10,000 feet on descent. Cabin crew on Starlink-equipped aircraft increasingly use the same link for payment terminals and electronic flight bag updates, which is part of why the business case closes even when the WiFi itself is given away to passengers.

FAA and EASA Certification Status [2026]

Starlink hardware cannot simply be installed on a certified aircraft. Any external antenna, radome, structural reinforcement and power feed constitutes a major alteration, and it must be covered by a Supplemental Type Certificate (STC) — an approval that says a specific modification is airworthy on a specific aircraft type.

What an STC actually certifies

An STC holder takes responsibility for three things under 14 CFR Part 25 for transport-category aircraft:

• Structural integrity: the radome, its doublers and the crown reinforcement must survive flight loads, bird strike and tens of thousands of pressurisation cycles.
• Electrical load analysis: the terminal’s power draw must not compromise existing systems under any load-shed condition.
• Electromagnetic compatibility: the transmitter must not interfere with navigation, communication or radar altimeter equipment — the item that turns certification into a multi-quarter exercise rather than a multi-week one.

Who holds the certificates, and how to verify it

SpaceX does not hold these certificates itself. FAA STCs for Starlink Aviation installations are held by partner avionics manufacturers, integrators and MRO organisations that engineer the installation for each airframe family, and the set of holders has expanded steadily since the first approvals in 2023.

Rather than repeat certificate numbers circulating in vendor marketing and trade coverage, the correct procedure — and the one an airworthiness inspector will expect — is to search the FAA’s Supplemental Type Certificate database at registry.faa.gov by make and model, then read the applicability list attached to the certificate itself. That document is the only authoritative statement of which holder covers which airframe, and it is updated as approvals expand. For non-US registries, the equivalent national authority database performs the same function. Treat any certificate number you read in an article, this one included, as a pointer rather than a source.

Europe: validation, not duplication

EASA does not generally re-certify from zero. Under the bilateral aviation safety agreement between the US and the EU, EASA validates FAA STCs, a process that is faster than original certification but still adds months and can attach additional conditions. That validation queue — not airline appetite — has been the practical rate limiter on European adoption, which is why Baltic and Gulf carriers moved ahead of most legacy EU flag carriers. The wider political and regulatory friction around SpaceX in Europe is a live story in its own right; see our coverage of Starlink in the European Union: Regulatory Landscape and National Rollouts.

One nuance operators routinely miss: an STC is type-specific and often serial-range specific. Approval on a Boeing 737-800 does not carry to a 737 MAX, and approval on an Airbus A321ceo does not automatically cover the A321neo. Fleet planners who assume otherwise lose a quarter to re-engineering.

Airline Partners and Rollout Timeline

The signal worth separating here is committed contract versus trial. A trial is a handful of aircraft and a press release; a commitment is a fleet-wide capital programme with MRO slots booked and hardware on order.

Two things stand out in the 2026 data. First, momentum is now concentrated in narrowbody domestic flying rather than premium long-haul, which is the inverse of how satellite connectivity historically diffused — legacy systems went onto flagship widebodies first because only those routes could justify the hardware. American’s programme alone, more than 500 aircraft from Q1 2027, is larger than the entire installed Starlink airline base of 2024. Second, the commitment curve has steepened sharply: 23 of the roughly 48 committed airlines signed during 2026.

For a US domestic passenger, the practical answer is that meaningful route-level coverage arrives through 2027 as the Southwest and American fleets convert, with JSX and Hawaiian already near-certain today.

How to Check Starlink Availability on Your Flight

Because equipment is allocated by tail number rather than by route, general Starlink availability claims are close to useless at the point of booking. A workable checking routine looks like this:

• Identify the registration, not just the flight number. Flight trackers such as Flightradar24 and FlightAware display the aircraft registration for a scheduled flight, usually firming up 12-24 hours before departure as the airline assigns equipment.
• Cross-check against the airline’s connectivity page. Carriers mid-retrofit increasingly publish which sub-fleets are converted; Hawaiian and JSX are effectively blanket yes, while United and Southwest remain sub-fleet dependent.
• Read the booking flow’s WiFi badge carefully. Several airlines now distinguish “fast, free WiFi” from generic “WiFi available” in seat maps and apps, which is the retrofit status leaking through the UI.
• Accept that equipment swaps happen. A tail substitution three hours before departure can silently downgrade you from 2Gbps-class Starlink to 2Ku. If the connection is business-critical, plan around it rather than on it.

Third-party tools that score a flight number’s Starlink probability exist precisely because airlines cannot yet guarantee equipment by route. That gap closes only when a carrier finishes a fleet, which is why the 2027 completion dates in the table above matter more to passengers than the 2026 announcement dates.

Performance at Altitude: Speed, Latency, and Reliability

Two generations of numbers are circulating, and conflating them is the most common error in aviation WiFi coverage.

• Early retrofits (2023-2024): user-submitted passenger speed tests clustered in the 40-100 Mbps per-aircraft range on lightly loaded cabins. These are crowd-sourced measurements taken over cabin WiFi, not certified link performance, and should be read as historical context rather than a current spec.
• Current installations (2026): systems rated at more than 2Gbps of total bandwidth per aircraft, reflecting newer terminals and far denser satellite capacity overhead.
• Latency: 20-40ms, against 600-800ms for GEO incumbents.
• Concurrency: on Aero-class terminals, roughly 20-50 simultaneous 4K streams is a reasonable working estimate; on 2Gbps-class installations, the binding constraint shifts off the satellite link entirely and onto the cabin’s wireless access points and the aircraft’s internal distribution.

Latency is the metric that changes behaviour. At 600ms, a video call is technically possible and socially unusable — every exchange collides. At 20-40ms, Zoom, Teams, SIP voice, remote desktop, VPN tunnels and even cloud gaming behave approximately like a mediocre home connection. That is the qualitative break, and it is why business aviation adopted before the airlines did: the people paying for heavy jets are the people whose hour is priced by the call they cannot take.

Weather and the altitude advantage

Rain fade — signal absorption by precipitation in Ku and Ka band — is the classic weakness of satellite links. An aircraft at 10,000 metres sits above the overwhelming majority of the troposphere’s water content, so the atmospheric column between terminal and satellite is thinner and drier than for any ground installation. Cabin service through weather that would degrade a rooftop dish is generally unaffected. The residual failure modes are different: steep bank angles that briefly compromise the beam geometry, and congestion when a dense corridor puts many equipped aircraft under the same satellite cell at the same time.

For context on how these figures compare with terrestrial and maritime satellite internet measurements from the same constellation, our ongoing testing coverage is collected in Starlink Real-World Performance: Speed Tests, Latency, and User Experience.

Starlink Price for Aviation: Three Cost Layers in One Table

Public commentary routinely mashes three separate cost layers together and then reports the confusion as a contradiction. There is a one-off capital cost, a recurring service cost, and a passenger-facing charge that may or may not exist. Here they are side by side.

Why the airline and private-jet monthly bands differ

The bands look inconsistent because they are quoted from opposite ends of the market. Airlines buy capacity as a fleet-wide volume commitment, negotiated privately and amortised across hundreds of tails, which pulls the effective per-aircraft rate down and caps the realistic top of the band well below retail. Business-aviation operators buy from a published tier card with no volume leverage, so the range extends higher — the $15,000 ceiling reflects an unlimited, priority-data plan for a single heavy jet, a product no airline would ever purchase one aircraft at a time. Same underlying capacity, two entirely different procurement mechanics.

The overlooked line item in both columns is the STC engineering itself. If an approved certificate already covers your type, you pay a fraction of the cost of being the launch customer for a new one. Operators of less common airframes should ask which certificate their MRO intends to work under, and confirm it on the FAA registry, before accepting any quote — the answer swings the total by six figures.

Starlink vs Gogo and Viasat: Competitive Comparison

On raw capability the comparison is not close: roughly an order of magnitude on speed and about twenty-fold on latency, plus the absence of dead zones over oceans where GEO beam coverage thins. The interesting question is therefore not whether Starlink wins on performance but why adoption is not instantaneous.

The answer is contractual. Legacy in-flight connectivity agreements typically run five to ten years, frequently with capacity minimums and early-termination penalties, and many were signed with hardware subsidies amortised across the term. An airline sitting three years into a seven-year Viasat deal cannot simply walk away without writing off the subsidy and paying to strip and replace equipment it has already installed. What 2026 shows is those contracts reaching natural expiry in clusters — which is precisely why so many commitments landed in a single year rather than spreading evenly since 2023. Delta remains the clearest counter-example: it has made no public Starlink commitment and continues with Viasat.

There is also a fleet-logic argument incumbents make honestly: dual-stack fleets are operationally awkward. Two connectivity systems means two spares pools, two training syllabi, two maintenance programmes and two sets of passenger expectations on the same route. Carriers that move tend to move decisively for that reason, which is why the 2026 announcements skew toward fleet-wide commitments rather than partial conversions.

Private Jet and Charter Installation Process

Business aviation adopted Starlink first because the decision sits with one owner rather than a board, and because a monthly service line item in the low thousands is trivial against the hourly operating cost of a heavy jet.

Step by step

• 1. Confirm STC applicability for your exact type and serial range. Early business-aviation approvals were publicised for airframes including the Gulfstream G650, Bombardier Global series and Dassault Falcon 8X, but the binding document is the certificate’s applicability list on the FAA registry — not the model family name in a press release.
• 2. Order hardware early. Reported lead times have run 8-12 weeks from order to delivery, which usually exceeds the wait for a maintenance slot, making the supply chain rather than the hangar your critical path.
• 3. Book a certified MRO. StandardAero, Duncan Aviation and Jet Aviation are among the facilities publicly offering these retrofits; their published programme descriptions put typical slot lengths in the two-to-four-week range for business jets.
• 4. Plan the downtime realistically. Crown structural work, radome fitting, power provisioning, cabin access point installation, then ground testing and at least one functional check flight.
• 5. File the paperwork. In the US, a major alteration requires FAA Form 337. Foreign-registered aircraft need the equivalent acceptance from their state of registry, and an aircraft on a non-US registry being modified under an FAA STC will need that validation recorded before return to service.
• 6. Combine it with a scheduled inspection. The single biggest cost saving available is folding the retrofit into an already-planned heavy check instead of buying separate downtime at the going hourly rate for a grounded aircraft.

Retrofit is the norm, not the exception — the overwhelming majority of Starlink-equipped business jets flying in 2026 were modified after delivery rather than line-fitted at the factory. Operators who move between jurisdictions should also understand how service terms, data tiers and roaming behave across borders, a topic we cover from the user side in Starlink for Digital Nomads: Portable Internet Across Borders.

Limitations and Future Developments

Three genuine constraints remain in 2026, and none of them are marketing problems.

High latitudes

Coverage thins above roughly 60 degrees north and south until enough Gen2 satellites occupy highly inclined orbits. That matters specifically for transpolar routings — Northern Europe to Northeast Asia, and some North America to Asia great-circle tracks — where service can degrade in the very sector where alternatives are also weakest, since GEO elevation angles collapse at those latitudes too.

Oceanic density

Ocean coverage depends on satellite density plus laser-link capacity rather than on ground stations. The North Atlantic track system is well served. Remote South Pacific and Southern Ocean sectors were the documented weak points through 2024-2025 and improve incrementally as launches continue.

Constellation evolution

The FCC’s partial authorisation of SpaceX’s Gen2 system, covering an additional 7,500 satellites, is the supply-side answer to both problems; the current grant conditions and any subsequent modifications are on file in the FCC’s satellite authorisation records. Larger Gen2 spacecraft carry more capacity per satellite and support direct-to-cell payloads alongside higher-bandwidth aviation beams, and Starship-class launch, as it matures, changes the mass-to-orbit economics of deploying them. The mechanics of that build-out — launch cadence, orbital shells and controlled deorbiting — are detailed in Starlink Satellite Constellation: Orbital Mechanics, Launch Cadence, and Deorbiting.

Inter-satellite laser links deserve particular emphasis because they are the feature that makes aviation viable at all. By routing traffic optically between spacecraft, the network stops depending on a gateway being within the satellite’s footprint — the same architectural property that lets terminals work where terrestrial infrastructure has been destroyed, as documented in Starlink in Disaster Relief: Hurricane and Conflict Zone Connectivity.

Competition and corporate structure

Eutelsat OneWeb sells LEO aviation capacity through distribution partners, and Amazon’s Project Kuiper has signed aviation agreements with JetBlue among its publicly named carriers, with service entry dates that should be checked against current FCC filings rather than launch announcements. On the corporate side, Starlink operates as a business unit of SpaceX, a private company; Starshield, the defence-oriented line, is contractually separate, and recurring reports of a Starlink spin-off or public listing remain unconfirmed as of September 2026.

The Bottom Line and Update Log

In 2026 Starlink has moved from novelty to default assumption in new in-flight connectivity procurement. The technical case — 20-40ms latency, gate-to-gate coverage, 2Gbps-class aircraft bandwidth — is settled, and it is the same satellite internet architecture that has already reset expectations on the ground and at sea. What remains unsettled is the schedule: roughly 48 committed airlines still have thousands of individual aircraft to convert, the largest programmes do not begin until Q1 2027, EASA validation continues to pace European fleets, and high-latitude and remote-ocean coverage depends on Gen2 deployment. If you are a passenger, check the tail number. If you are an operator, check the STC applicability list on the FAA registry before you check the price.

Last updated: September 2026, with Q3 airline rollout confirmations.

Changelog: Added United Airlines trial expansion (July 2026); added American Airlines 500-aircraft commitment announced May 2026 and Southwest’s early-2026 pivot; corrected airBaltic timeline from 2024 to 2025; June 2026 correction — removed a prior reference to a Delta Air Lines commitment, as Delta remains with Viasat per its Q2 2026 earnings call; September 2026 revision — replaced previously cited STC certificate numbers and holder attributions with a pointer to the live FAA registry after they could not be independently verified, and reframed 2023-2024 speed figures as crowd-sourced user reports rather than measured results.

Sourcing and verification notes: FAA Supplemental Type Certificate database (registry.faa.gov) for certificate holders and applicability lists; FAA Form 337 requirements under 14 CFR Part 43 Appendix A and airworthiness standards under 14 CFR Part 25; EASA type certificate and validation databases for European approvals; FCC satellite authorisation records for the SpaceX Gen2 constellation; SpaceX and Starlink Aviation published service and plan documentation for hardware specifications and retail pricing tiers; airline press releases and investor disclosures from JSX, Hawaiian Airlines, American Airlines, Southwest Airlines and Viasat; published MRO programme materials from StandardAero, Duncan Aviation and Jet Aviation for installation timelines. Airline-side monthly service costs are industry estimates, not disclosed contract terms, and are labelled as such. Speed figures attributed to 2023-2024 are historical, user-submitted measurements and should not be read as current or guaranteed performance. Readers making procurement decisions should verify certificate applicability and current pricing directly with the FAA registry and SpaceX rather than relying on any secondary source, including this article.

Frequently Asked Questions

Is Starlink available on commercial flights yet?

Yes. Starlink has been flying commercially since JSX switched on its first aircraft in Q1 2023, and by September 2026 roughly 48 airlines worldwide have committed to the system, 23 of them signing during 2026 alone. Hawaiian Airlines, Qatar Airways, airBaltic, Air New Zealand and United all have aircraft in service or in active installation. The largest programmes — American Airlines with more than 500 narrowbodies and Southwest with over 300 — are still mid-rollout, so Starlink availability depends on the specific tail number rather than the airline. Check the aircraft registration on a flight tracker before you book if connectivity matters to your trip.

How much does Starlink cost on a private jet?

Published business-aviation plans sit in a roughly $2,500 to $15,000 per month band, scaling with the data tier from around 150GB to unlimited. Light jets typically sit near the $2,500 entry tier; heavy jets on unlimited, priority-data plans occupy the upper half of the range. Those are service fees only. The terminal, radome, wiring and STC engineering add roughly $150,000 to $250,000 per aircraft as a one-off, and EU-registered operators should expect VAT on top of quoted figures. Airlines pay less per aircraft than this retail band because they negotiate fleet-wide volume contracts.

Is Starlink Aviation the same service as home Starlink internet?

Same constellation and same satellite internet backbone, different terminal, different certification and a very different price. Residential Starlink uses an unpressurised ground terminal on a consumer plan costing tens of dollars a month. Starlink Aviation uses the Aero terminal — a hardened, low-profile phased array certified for flight loads, pressurisation cycles and electromagnetic compatibility with cockpit avionics — under a managed service contract costing thousands of dollars a month. The satellites overhead are the same; everything between them and the user is not.

Can you stream video on Starlink during a flight?

Yes, and unlike legacy in-flight WiFi, streaming is not the edge case it was. Latency of roughly 20-40ms is low enough for synchronous video calls, cloud gaming and VPN work, not just buffered playback. On 2026-generation installations delivering more than 2Gbps of total aircraft bandwidth, dozens of simultaneous high-definition or 4K streams are realistic. On earlier retrofits, where user-submitted speed tests clustered around 40-100 Mbps per aircraft, streaming works but degrades noticeably on a full cabin.

What is the difference between Starlink and Gogo in-flight WiFi?

Gogo 2Ku relies on geostationary satellites orbiting at 35,786km, which produces 600-800ms round-trip latency and typical passenger throughput of about 3-9 Mbps. Starlink works from roughly 550km, cutting latency to 20-40ms and raising aircraft-level bandwidth by an order of magnitude or more. The commercial model differs too: Gogo has long sold $10-20 per-flight sessions, while Starlink carriers such as Hawaiian give access away free or bundle it into loyalty programmes.

How long does it take to install Starlink on an aircraft?

Budget 2-4 weeks of aircraft downtime at a certified MRO such as StandardAero, Duncan Aviation or Jet Aviation, assuming an STC already exists for your type. Hardware lead times have been reported at 8-12 weeks from order to delivery, so the calendar, not the wrench time, is usually the constraint. Paperwork matters as much as the install: a major alteration needs FAA Form 337 in the US, and foreign-registered aircraft need equivalent filings with their own authority.

Does Starlink work on flights over the ocean?

Over most oceanic routes, yes. Inter-satellite laser links on v1.5 and later satellites let traffic hop between spacecraft instead of needing a ground station in view, which is what makes mid-Atlantic and mid-Pacific coverage possible at all. The North Atlantic track system is well served. The weakest links have historically been remote South Pacific and Southern Ocean sectors and high-latitude transpolar routes above roughly 60 degrees, where satellite density and orbital inclination still limit throughput.