China is building two broadband satellite constellations in parallel, with more than 20,000 spacecraft planned between them.
Guowang, also known as SatNet, is operated by China Satellite Network Group, a central state-owned enterprise created for the project in 2021. Qianfan, or Spacesail, is operated by Shanghai Spacecom Satellite Technology and backed by the Shanghai municipal government.
Neither operator has a public English-language website, so the deployment figures below are drawn from SpaceNews launch reporting and public orbital catalogs. Both constellations are intended to do the same job as Starlink, but both are running behind the schedules set out in their own filings.
The satellite counts are the number to watch. By mid-2026, roughly 190 Guowang satellites had launched, while Qianfan passed 200 in June 2026. Starlink, by comparison, operates thousands.
Guowang and Qianfan Compared
| Measure | Guowang (SatNet) | Qianfan (Spacesail) |
|---|---|---|
| Operator | China Satellite Network Group, central state-owned | Shanghai Spacecom Satellite Technology, trading as Spacesail |
| Other names | SatNet, GW constellation | Thousand Sails, G60 Starlink Plan |
| Satellites launched | About 190 by mid-2026 | Passed 200 in orbit, June 2026 |
| Planned size | Around 13,000 | 15,000 filed |
| 2026 deployment target | 310 satellites | 324 satellites |
| Stated scale-up | 900 in 2027, 3,600 a year from 2028 | 324 in 2027, several thousand a year from 2028 |
| First launch | December 2024 | August 2024 |
| Market focus | National infrastructure, government and military users assumed | Commercial, including agreements abroad |
Counts change with nearly every launch. Treat the numbers above as a September 2026 snapshot and check a live catalog such as CelesTrak before quoting them.
Both Are Behind Their Own Schedules
The gap between plan and delivery is the clearest fact about both constellations. Guowang plans 3,600 satellites a year from 2028 and launched roughly 190 in its first eighteen months. Qianfan has filed for 15,000 and has a few hundred up.
Qianfan’s deployment also stopped for months. Launches paused after early satellites suffered thruster and gyroscope failures in orbit, and resumed in April 2026 after corrective work, according to SpaceNews. Six launches followed in quick succession once the pause lifted.
Launch Capacity Is the Constraint
Filling two constellations of 10,000-plus satellites requires a launch rate China does not yet have. Each Long March flight carries a modest batch and the rocket is then discarded, so the arithmetic does not close at any expendable flight rate the country has achieved.
China flew a national record 92 orbital launches in 2025 and targeted roughly 140 for 2026. Even at 140 flights a year, with only a share of them carrying constellation batches, the deployment schedules above stay out of reach.
That is the demand behind China’s push into reusable rockets. Two boosters came home in 2026, the Long March 10B in a sea catch and Zhuque-3 on legs, and neither has flown again. The programs are covered on the China reusable rockets page.
The satellites themselves have gotten cheaper in parallel. Qianfan has moved to flat-panel satellites built for stacking, the same manufacturing shift that made Starlink batches practical, and CGTN has reported the redesign cutting per-satellite cost by more than 90 percent.
What Each Constellation Serves
Guowang is treated as national infrastructure. It is run by a central state-owned enterprise created for the purpose, its filings reserve large amounts of spectrum, and government and military communications are assumed users. A sovereign broadband network that does not depend on a foreign operator has obvious value to a state that treats BeiDou navigation the same way.
Qianfan has behaved more like a commercial venture. Spacesail has pursued service agreements in countries including Brazil and Kazakhstan, positioning the network as an alternative for governments that want broadband without depending on an American provider.
Neither is available in most Western markets. Operating a satellite broadband service requires spectrum and landing rights from each country’s regulator, and Chinese operators have not obtained those in North America or most of Europe.
The Other Chinese Constellations
Guowang and Qianfan get the attention because of their size, and they are not the only Chinese networks in low Earth orbit. Several smaller constellations serve narrower jobs.
- Honghu-3 is a commercial broadband constellation with its own filings, and Honghu satellites have flown on Zhuque-3.
- Jilin-1 is a large commercial Earth-observation constellation operated by Chang Guang Satellite Technology, imaging rather than communications.
- Yaogan is a long-running series of military reconnaissance satellites, launched steadily for years and described officially as remote sensing.
- BeiDou is the navigation network, in medium Earth and geostationary orbits rather than low Earth orbit, and complete since 2020.
Adding all of them together produces figures like “China has hundreds of satellites”. They do different jobs on different orbits, and a broadband count and a navigation count should not be added.
How They Compare to Starlink
Starlink is years ahead on every operational measure: satellites in orbit, subscribers, coverage, and launches per year. It also has an advantage neither Chinese constellation has, which is a reusable rocket owned by the same company that builds the satellites.
That vertical integration is the structural difference. SpaceX launches Starlink on its own reflown Falcon 9 boosters at a marginal cost it controls. Guowang and Qianfan buy launches from CASC and from commercial Chinese firms, all of which currently fly expendable rockets.
China competes on equal terms for spectrum. Filings with the International Telecommunication Union (ITU) are first-come, and both Chinese constellations filed for large allocations early. Orbital shells and spectrum are finite, and a constellation that never fully deploys can still hold its claim.
China Runs Two Constellations as a Hedge
Running two megaconstellations at once looks like duplication and works as a hedge. Guowang answers to the central government and Qianfan to Shanghai, which means two funding streams, two engineering teams, and two sets of supplier relationships chasing the same capability.
China has used this pattern before. Two design academies inside CASC build overlapping Long March variants, and six private companies are developing reusable rockets in parallel. Competition between the two sets a pace that a single program would not.
The cost is spectrum and orbital shells claimed twice over, plus satellites built to two incompatible designs. For an operator that is waste. For a government treating broadband as infrastructure, it is redundancy.
Who Should Track These and Who Should Not
Anyone modeling global launch demand, orbital congestion, or satellite broadband competition has reason to follow both constellations, because 20,000 planned satellites is a material share of everything humans intend to put in low Earth orbit this decade.
Consumers looking for satellite internet service should look elsewhere. Neither network sells to individuals in most countries, and the best satellite internet providers page covers what is available today.
What Would Change the Picture
A sustained launch rate is the number to watch. If either constellation strings together enough flights to hit its stated annual target, the schedules stop being filings and start being a deployment curve.
A reflown Chinese booster would change it faster. Reuse is the only lever that raises China’s flight count enough to matter at 3,600 satellites a year. To check current satellite counts before citing them, consult a live catalog such as the one maintained by CelesTrak rather than a news figure.