When a technology is still in its infancy, the “pioneer” often holds the greatest value. The first to launch, orbit, or return sets a precedent, acting as a milestone that can define a company, a technical approach, or even an entire phase of an industry.

By December 2025, this milestone stood between ZQ-3 and LM-12A. Two liquid oxygen-methane rockets completed their maiden flights just 20 days apart, with their second stages successfully reaching predetermined orbits. However, both failed to stabilize their first stages at recovery sites. The world watched to see which would be the first to add the return link to their next flight, becoming China’s inaugural rocket to achieve orbital-class first-stage recovery.

The answer came sooner than expected—from an unexpected contender.

On July 10, 2026, the LM-10B lifted off from the Hainan Commercial Space Launch Site. Its first stage returned to the sea and was captured by a net system aboard the “Pathfinder” recovery vessel. This marked China’s first controlled recovery of a launch vehicle’s first stage and the world’s first successful net capture of a launch vehicle at sea. According to the China Aerospace Science and Technology Corporation (CASC), the recovered first stage will undergo inspection, maintenance, and retesting, with plans for a reused flight before year’s end.

The first milestone has been achieved, yet ZQ-3 and LM-12A must still take flight.

ZQ-3 Y2 completed static firing on June 29, while LM-12A Y2 is scheduled for subsequent recovery tests. Specific dates may shift, but when they return to the launch pad, the race for records will take a backseat. The focus will shift to whether domestic landing leg methods can achieve first-stage recovery and how far private rockets and commercial propulsion are from stable reuse.

01

After LM-10B’s recovery, China has crossed a significant threshold, but reusability remains unproven.

LM-10B, LM-12A, and ZQ-3 all aim to recover their first stages, but their terminal recovery methods differ substantially. LM-10B relies on engines, grid fins, and control systems for return and deceleration, ultimately captured by a flexible net system on a maritime recovery vessel. In contrast, LM-12A and ZQ-3 use self-contained landing legs to control speed, attitude, and landing point during terminal ignition, ultimately standing on land-based recovery sites.

Thus, the so-called “debate between vertical and net recovery” is somewhat misleading; LM-10B also undergoes powered deceleration and vertical return. The distinction lies in whether the structural, load, and support capabilities required for terminal recovery remain primarily on the rocket or are partially transferred to a maritime platform.

Net capture reduces the rocket’s reliance on landing legs and related load-bearing structures, with the flexible net absorbing some impact. However, it demands a massive maritime support system. The “Pathfinder” is 144 meters long, 50 meters wide, with a displacement of approximately 25,000 tons, backed by resources for towing, surveillance, tracking, port operations, and rocket transportation. While net capture reduces weight on the rocket, it shifts another portion of the cost to the sea.

The landing leg approach retains support structures, deployment mechanisms, and touchdown loads on the rocket, imposing stricter requirements on terminal ignition, thrust control, and landing stability. However, as demonstrated by the Falcon 9, it supports high-frequency recovery and relauches, with the flexibility to choose land or sea platforms based on mission needs.

Net capture has completed one real mission, while the domestic landing leg route still lacks successful landings after orbital launches. If either ZQ-3 or LM-12A stabilizes, China will simultaneously possess two orbitally validated first-stage recovery methods.

02

The maiden flight tests the overall design in a real-world environment, while the second flight (Y2) validates fault localization and improvements from the previous flight.

Flight data must be translated into failure mechanisms, leading to design modifications, ground replications, software adjustments, configuration management, and batch manufacturing. The closure report provides conclusions, which the next flight verifies.

Currently, public information on the two failed recoveries is asymmetrical. LandSpace revealed that ZQ-3 Y1 achieved high-precision guidance to the recovery site during re-entry ignition and aerodynamic gliding phases but experienced abnormal combustion several kilometers above the ground, failing to achieve a soft landing. Less information is available for LM-12A. The China National Space Administration (CNSA) only confirmed that the first stage failed to recover successfully and that the development team would conduct a comprehensive review and technical closure without disclosing which system malfunctioned.

Under these circumstances, claims like “ZQ-3 was only one step away” or “LM-12A’s first retro-propulsion ignition failed” should not be treated as definitive conclusions. Flight footage provides clues but cannot replace telemetry data, nor can it rank the difficulty of the two failures.

Only if Y2 reliably reproduces the previously completed flight phases and demonstrates continuous operation of the terminal ignition, guidance control, and landing systems can it prove that the closure measures withstood real-world conditions. If abnormalities recur in the same or similar phases, the previous fault localization and improvement effectiveness must be re-examined.

03

LM-10B’s pioneering completion of recovery has not diminished the significance of private rockets developing reusable technology but has pushed industry evaluation criteria forward.

In recent years, liquid oxygen-methane propulsion, grid fins, landing legs, and “designed for multiple reuses” have formed an attractive technological narrative. After the national team achieved real recovery, these terms have lost their scarcity. Solutions and design specifications remain valuable, but external attention will increasingly shift toward actual recovery, vehicle inspection, and reused flights.

Currently, only LM-10B has completed one orbital-class first-stage recovery in China, insufficient for comparing recovery frequency and turnaround efficiency among various entities. Questions like how many times a first stage can fly, how long inspections take, and whether reuse can lower prices remain unanswered with industry-wide comparable records.

For LandSpace, ZQ-3 Y2 faces stronger external scrutiny. It is China’s first private rocket attempting orbital-class first-stage recovery and the cornerstone of LandSpace’s heavy-lift launch services. Before LM-10B’s success, it represented China’s closest attempt at orbital-class first-stage recovery; afterward, it must prove that a privately developed rocket can complete the engineering chain of orbital launch and first-stage recovery.

Successful recovery is not a mandatory condition for commercial rocket enterprises to meet the fifth set of listing standards on the STAR Market. The relevant guidelines require phased achievements, namely, the first successful orbital insertion of a payload by a medium-to-heavy launch vehicle using reusable technology, which ZQ-3 Y1 has already achieved.

However, the capital market will still use Y2 to assess how much more time and funding LandSpace needs to transition from orbital insertion to recovery, reuse, and stable services.

In the past, recoverable designs could expand valuation prospects; after the national team’s achievement, prolonged unfulfilled recovery goals will erode such imagination.

04

LM-12A Y2 also carries an industrial trial easily overshadowed by recovery imagery.

According to public reports, LM-12A’s first stage uses seven Jiuzhou Yunjian Longyun series liquid oxygen-methane engines, with the Eighth Academy of Aerospace Science and Technology Group (CASC) responsible for overall development. Commercial propulsion products have thus entered the main propulsion chain of Long March rockets, beginning to challenge the long-standing in-house supply system within academies and institutes by accepting market-oriented core suppliers.

The main engine directly affects the rocket’s thrust, flight control, and mission reliability. After integration into the Long March system, it must undergo interface control, batch acceptance, quality traceability, multi-engine paralleling, and full-vehicle environmental verification; the overall developer must also establish mechanisms for data sharing, collaborative modifications, and responsibility division with external suppliers.

If this model withstands consecutive flights, China may finally see truly independent commercial engine enterprises supplying multiple rocket models. With more models and sustained orders, engine R&D, testing, and production line investments can be amortized over scale, and overall developers won’t need to replicate a propulsion system for each new model.

Before the official closure report is released, LM-12A Y1’s recovery failure cannot be attributed to the Longyun engine. First-stage return involves multiple systems, including propellant management, engine restart, navigation guidance, attitude control, thermal protection, and structure. Unverified single-point attributions are neither accurate nor helpful in diagnosing the issue.

ZQ-3 adopts a different organizational approach. LandSpace independently develops engines, vehicle bodies, and recovery systems while constructing dedicated launch pads, keeping more critical links in-house. Thus, the comparison between the two Y2 missions involves not only recovery technologies but also two industrial organizational models: external supply chain collaboration versus vertical integration.

05

Recovery transforms from a flight issue to a cost issue after the first stage lands.

Engineers must assess the condition of the vehicle structure, thermal protection, engines, and avionics systems, determining which components can be reused, which require maintenance or replacement, and thereby defining inspection scope, overhaul workload, and turnaround time.

Recovery saves the cost of manufacturing a new first stage, but costs for recovery vessels or sites, testing equipment, maintenance personnel, overhaul cycles, and the payload capacity loss due to recovery must also be accounted for. If each returned first stage requires extensive disassembly and repairs, reuse may merely shift manufacturing costs to overhaul costs.

LM-10B has currently achieved recovery, but relauches remain unverified. Even if ZQ-3 and LM-12A land in their next flights, they will only have brought back first stages that have undergone real orbital missions. Whether they can fly again and the maintenance costs involved will only be answered after inspection and relauches.

Thus, after LM-10B, ZQ-3 and LM-12A are no longer competing to be “China’s second.” If ZQ-3 stabilizes, it will mark the first time a private enterprise has achieved first-stage recovery after orbital launch. If LM-12A stabilizes, it will add crucial flight experience to the collaborative model between national team developers and commercial propulsion providers. What both must accomplish is the orbitally validated recovery test that domestic landing leg methods have yet to complete.

One net has caught China’s first success. Next, we’ll see whether the two first stages that previously missed their landings can stand firmly on their own landing legs.