In less than 16 hours, two rockets from different private commercial spaceflight companies sent 18 Qianfan satellites into space.
The first rocket came from LandSpace. At 14:26 on September 15, the improved Zhuque-2 Y7 from LandSpace lifted off from the Dongfeng Commercial Space Innovation Pilot Zone, sending 10 Qianfan polar orbit Batch 19 satellites into their predetermined orbits. This marks the first time a private commercial spaceflight enterprise in China has undertaken the constellation networking task for a large-scale satellite internet, signaling that China's private commercial spaceflight has officially entered a new stage of batch networking.
15 hours and 34 minutes later, at 06:00 on September 16, Orienspace's Gravity-1 Y3 lifted off from the eastern sea area of Shanghai, sending 8 Qianfan networking satellites and 1 EUHT technology test satellite into an orbit of approximately 800 kilometers. This launch was also the first time
globally to execute a LEO broadband internet constellation networking mission via mobile sea-based launch.
Five years ago, the most critical issue for China's private rockets was still whether they could successfully reach orbit. Now, as more and more companies have crossed this threshold, the question is turning into another phrase more akin to manufacturing: after the rockets are built, can they be delivered safely and steadily, one after another?
But delivery is just the surface. The real question is who will pay for this long-termism of high-frequency delivery.
A Major Client Arrives for Private Spaceflight
Commercial rockets have always faced a fundamental contradiction. Developing a rocket requires several years and massive investment, but there are not many clients who can stably purchase launch services. Remote sensing satellites, experimental satellites, and small constellations can bring in orders, but it is hard for a rocket company to plan its capacity at a pace of over a dozen or even dozens of launches per year based on them.
Large LEO constellations like the Qianfan constellation are primarily changing this demand structure. Gao Tianwei, a commercial spaceflight expert, believes that large constellation orders are currently the "sexiest" source of orders.
This "sexiness" comes not only from the order volume but more from the fact that demand is becoming continuous and predictable. It means that satellite enterprises no longer need a specially customized rocket that is used once and done; instead, they need a launch capacity product that can be repeatedly purchased.
In 2025, Yuanxin Satellite (the operator of the Qianfan constellation) allocated CNY 1.336 billion to procure 7 launch services, planning to launch 94 satellites, and directly divided the procurement specifications into "10 satellites per rocket" and "18 satellites per rocket". The project adopts "in-orbit delivery", with clear regulations on the rocket's target orbit, payload capacity, delivery time, and even the maximum price per kilogram. The maximum price limit for 10 satellites per rocket is CNY 55,000 per kilogram, and for 18 satellites per rocket, it is further reduced to CNY 50,000 per kilogram.
Gao Tianwei told KEGONG that the current price for domestic commercial launches is roughly around CNY 50,000 per kilogram. Calculated at this price, if a rocket undertakes a constellation networking task of 2.8 to 4.8 tons, the theoretical revenue for a single launch service would be approximately CNY 140 million to CNY 240 million. However, Gao also pointed out that the actual transaction price generally comes with a significant discount.
Calculated based on corporate prospectuses and disclosed cost data for similar rockets, the single launch cost of LandSpace's Zhuque-2 is about CNY 130 million, and the single launch cost of CAS Space's Gravity-1 is about CNY 100 million. Combining the above data, commercial spaceflight companies may have just touched the edge of "being able to cover costs", and are still far from stable profitability, especially filling the massive upfront R&D investments.
But for current commercial rocket companies, whether a single launch can make money is not the top priority. In explaining why its launch service costs remain high, CAS Space's prospectus mentioned that due to the limited production of rockets, economies of scale have not yet been formed, and fixed costs such as factories and equipment can only be amortized over a small number of missions; at the same time, many upstream components still rely on customized production, making it difficult for suppliers to lower their quotes through mass manufacturing.
Therefore, only by securing sufficiently stable orders can commercial spaceflight companies transition engines, rocket bodies, and final assembly from small-batch customization to mass production. Only then can the fixed costs of factories and launch facilities be diluted by more missions, making subsequent continuous cost reduction possible.
This is precisely where large constellation orders are truly "sexy". What large constellations really change in the commercial rocket industry is not just bringing more orders, but providing a realistic premise for scaled production for the first time.
Public information shows that commercial rocket enterprises such as LandSpace and CAS Space have all entered the procurement system related to the Qianfan constellation. According to Yuanxin Satellite's plan, the goal is to achieve a cumulative 324 satellites in orbit this year, and currently, 68 satellites are still needed to reach this target. Gao Tianwei estimates that, based on the launch frequency and networking pace of the Qianfan constellation since the beginning of this year, the remaining 68 satellites have entered the executable range, and there is a possibility of completing the target within the year.
Beyond Qianfan, another 10,000-satellite-class LEO satellite internet in China is also continuously networking. As more and more commercial spaceflight companies possess stable launch capabilities, the commercial spaceflight industry is also ushering in an order release period and a large-scale fulfillment window.
But after the orders arrive, what is truly tense may not be the production line, but the clock of capital.
In the first quarter of 2026, China's commercial spaceflight completed 46 financing rounds, totaling CNY 8.02 billion, a massive year-on-year increase of 4.6 times. More than a dozen companies are queuing for IPOs, and 2026 is regarded as the year of concentrated explosion for commercial spaceflight IPOs. The clock of capital is a "7-year exit", but the clock for rocket mass production is a "10-year ramp-up". When investors evaluate companies using "how many launches can be delivered this year", companies have to walk a tightrope between "delivering on time" and "process refinement".
This is not to criticize capital for being short-sighted. Rather, it is to remind us of a fact: the networking demand of Qianfan is a testing ground for "mission-oriented innovation", but much of the money pouring in carries KPIs from the "industrialization stage". Launch windows are constrained by orbital planes and networking schedules; missing one launch could affect the subsequent overall situation. The window of capital is constrained by the fund's duration; missing one round could affect the life and death of the enterprise. The desynchronization of these two clocks is the most hidden pressure behind the "competition on delivery".
After the Orders Arrive, the Rules of the Game Must Also Change
In July this year, KEGONG summarized the above supply and demand relationship as "satellites dictating rockets". Rocket companies that take the lead in achieving high-frequency fulfillment will not only find it easier to get orders, but may also gain greater bargaining power in subsequent pricing and service models. Currently, this trend is becoming more concrete.
The Zhuque-2 series has previously completed multiple orbit-reaching missions, but LandSpace still refers to this Y7 as the first time the Zhuque-2 series has executed a scaled satellite internet constellation networking mission.
The difference lies in the two words "networking". When launching a single independent satellite, the rocket only needs to deliver the payload to the designated orbit; when carrying ten formal networking satellites at once, it must simultaneously handle multi-satellite adaptation, separation sequence, orbital deployment, satellite-rocket interfaces, and mission windows.
Orienspace also faced a similar situation this time. Gravity-1 simultaneously carried stacked satellite combinations and flat-panel satellites, which also increased the difficulty of multiple separation controls.
This means that for commercial spaceflight companies, large-scale, high-frequency delivery has just begun, and they already need to face higher requirements.
In the past, the most direct standard to judge whether a commercial rocket is mature was "whether it can successfully reach orbit". But for a constellation operator that deploys orbital planes according to a set schedule, mere success is not enough; it also requires the rocket to succeed exactly when needed.
Gao Tianwei told KEGONG that the launch windows for large constellations are constrained by orbital planes, satellite production batches, and the overall networking schedule. If a rocket misses the window due to testing, troubleshooting, or other reasons, the impact will not be limited to this single mission; it could also disrupt subsequent satellite production, transportation, orbital deployment, and launch applications.
At this stage, commercial rocket orbit-reaching launches still implement per-mission licensing. Taking the commercial rocket licensing guidelines published by the Beijing Municipal Commission of Science, Technology and Industry for National Defense as an example, commercial rocket enterprises executing orbit-reaching missions "must apply for launch permits on a per-mission basis", which involves multiple contents such as flight safety, rocket factory release, drop zone safety, TT&C schemes, orbital parameters, radio frequencies, and insurance.
These procedures themselves are an indispensable part of spaceflight safety management, but when commercial rockets advance from a few launches a year to over a dozen or even dozens, the nature of the problem will also change: rocket production lines can be mass-produced, launch sites can add workstations, and the management and mission organization systems also need to adapt to higher turnover frequencies. This poses higher requirements for the delivery capabilities of commercial spaceflight companies.
Whether a company chooses liquid oxygen-methane or liquid oxygen-kerosene, what engine scheme is adopted, or even whether the first stage can be recovered, are not the only concerns for satellite companies during the procurement stage. The technical route must ultimately fall on a few key indicators: can this rocket be built on time, can it be launched on time, and can it accurately deliver the satellites to the positions required by the customers.
Of course, for commercial spaceflight companies, especially private ones, cost remains a must-consider issue, as it is the "life and death line" that determines how far they can go.
The Best Order Window May Also Be an Elimination Race
But as we are encouraged by the current "10,000-satellite networking" boom, and the gradually forming commercial closed loop of "increased constellation orders - high-frequency rocket launches - scaled cost reduction - further demand release", a question also arises: how long can this launch capacity demand last?
Gao Tianwei's answer is: perhaps not as long as imagined.
If simply calculated according to the current logic, Qianfan needs over 10,000 satellites, other constellations also plan for over 10,000, and superimposed with a constellation renewal cycle of about 7 years, then in the future, it seems that more and more rockets will always be needed.
But SpaceX is providing another answer.
The next-generation Starlink V3 single satellite downlink capacity is designed to reach 1 Tbps, which is an order of magnitude ten times that of the current generation of satellites; SpaceX estimates that a single Starship can deploy about 60 V3 satellites at once, and the downlink capacity added to Starlink by a single mission can reach about 20 times that of the current Falcon 9 launching V2 satellites.
In the future, when evaluating a constellation, simply counting "how many satellites there are" may become increasingly meaningless. More important indicators will become: how much capacity a single satellite has, how many tons of payload a single launch can deliver, and how much communication capacity a single rocket can add to the network.
This is consistent with the contradiction that has already emerged between domestic satellite and rocket parties: perhaps satellites do not need to exhaust the declared quantity to meet performance requirements.
From the perspective of satellite performance and R&D costs, constellation operators naturally hope to make single satellites larger and more powerful; rockets, on the other hand, hope to improve single deployment efficiency within the limits of the fairing and launch capacity. The coordination between satellite size, weight, interfaces, and launch processes will increasingly and directly determine the price of launch services.
Thus, a seemingly contradictory result may emerge in the future: satellites are getting heavier and heavier, the total tonnage of space transportation continues to increase, but the required number of satellites and launch frequencies may not grow synchronously at the speed of today's "10,000-satellite plan".
This is not good news for commercial rocket companies. During the networking period, customers lack launch capacity, and Zhuque-2, Gravity-1, the Long March series, and more new models may all get orders. Once the network deployment is finished and the market enters the stage of expansion, network replenishment, and satellite renewal, customers will recalculate a different account: who can load more satellites per rocket, who is cheaper per kilogram, who can provide more certain windows in a year, and whose rockets can still maintain reliability after reuse.
Once economies of scale are formed, the market is prone to positive feedback. The more launches, the lower the cost; the lower the cost, the more orders; the more orders will further promote the expansion of supply chains, factories, and launch sites around a few models.
Only enterprises that can take the lead in utilizing the demand window provided by this round of large-scale constellation networking, and run out the curves of reliability, capacity, launch frequency, and cost, are more likely to secure a place in the commercial spaceflight field in the future. Currently, this window period is not long, and the future commercial spaceflight market may not need as many companies as there are today.
This pressure comes not only from the demand side. The threshold on the supply side is also rapidly rising. In the past, private rockets mainly faced competition with each other. Now, the "national team" has also begun to directly participate in large constellation launches with commercialized products.
In June this year, the Long March 12B, developed by the "national team" commercial spaceflight company China Rocket, successfully completed its maiden flight. In terms of product form and performance, the Long March 12B is very close to the Falcon 9. It is inherently designed for large-scale internet constellations, with a LEO capacity of about 20 tons. The first stage uses 9 liquid oxygen-kerosene engines and continues to iterate in the direction of reusability.
In July, the Long March 10B, also belonging to the "national team", successfully conducted China's first launch vehicle orbit-reaching recovery test. The Long March 10B is also a medium-to-large commercial rocket with a maximum capacity of the 20-ton class, and the success of its recovery and even future reuse will endow the model with a huge cost advantage.
Gao Tianwei believes that the commercial rockets from China Rocket have actually raised the threshold for future competition in private commercial spaceflight. Launch capacity, reliability, and cost are the three key indicators for commercial rockets. Currently, the advantages of the "national team" in reliability and launch capacity are hard to shake. Private rockets can almost only focus on the cost side, while reliability and launch capacity cannot be too far behind; otherwise, it will be difficult for private rockets to secure a stable position in the large constellation launch market in the future.
Of course, from the perspective of satellites, China's commercial spaceflight cannot only catch up within SpaceX's coordinate system. SpaceX has defined the game rules of "single satellite capacity × launch tonnage". If China only benchmarks on "how many satellites per rocket" and "how much money per kilogram", it is just eating the wake—closely following the leader to reduce drag, but forever losing the ability to define the direction.
The true differentiation of Qianfan lies in its ground service capabilities of "integration of communication, navigation, and remote sensing". Satellite, navigation, and remote sensing data collaborate on a single network to provide turnkey solutions for specific industries. This is not a technical fantasy; the remote sensing ecosystem of Chang Guang Satellite has already proven the feasibility of this "chain leader logic". When SpaceX's Starlink covers every person in the world, China's constellation network can go to cover every piece of farmland in the world that needs to be remembered, every fishing boat that needs to be positioned, and every power transmission line that needs to be monitored. The two networks grow in parallel, neither subordinate to the other.
The future elimination race is about "who can define the standard of 'delivery'"—whether it is delivering a single satellite, or delivering a set of capabilities.