The construction pace of new energy vehicle charging networks has accelerated sharply this year, with station counts refreshing on a nearly monthly basis.

On August 28, 2026, Byd Company Limited (002594.SZ) unveiled its 10,000th flash-charging station in Shenzhen. According to its roadmap, the company plans to have 20,000 flash-charging stations operational by year-end, meaning the station count will double over the next four months.

CATL's subsidiary Times EV Service is also speeding up the deployment of battery-swap stations this year. Close sources indicate that after recently announcing a plan to build 100 stations in a single city in Kunming, it will roll out similar "100-station per city" projects in several other cities. Its goal is to surpass 3,000 swap stations nationwide by the end of the year.

Beyond these two new energy heavyweights, automakers that have spent years building their own networks continue to expand. Competition in the charging and swapping network space has clearly intensified this year.

The latest acceleration was driven by the speed gap between vehicle-side technology and station-side infrastructure. 800V architectures, high-rate batteries, and fast-charging platforms have entered more models, yet as of late July, the average single-gun power of public charging facilities nationwide remained around 49.97 kW. The same vehicle plugging into different stations can experience actual charging power differences of severalfold. When station capabilities lag behind, the peak power figures touted by automakers struggle to translate into real user experience and purchase reasons.

The next wave of sales growth for new energy vehicles still has to be won from the fuel car market. High-voltage platforms compress charging times, solving only whether a vehicle can fast charge; whether cities, highways, and frigid regions have dense and stable enough networks determines if fuel car users can let go of the certainty that refueling provides. Vehicles have already sprinted ahead, and the gaps left in the charging network are turning into the entry point for the next round of sales competition.

That is precisely where the current window lies. Once a leading network achieves density, late entrants will have to re-fight for locations, traffic, and vehicle model compatibility even if they can source the equipment.

The recent flurry of activity from companies like Byd Company Limited (002594.SZ) and CATL (03750.HK) has pushed this still-unsettled phase into the spotlight. The locations locked in first and the vehicle models integrated early will both lower the cost of the next expansion round.

A charging war has now begun. Competition in the auto industry is extending from building a good car to organizing an energy network.

Grabbing Stations Is Also Grabbing Vehicles

Sources from Byd Company Limited (002594.SZ) indicate that its latest growth has shifted toward densifying urban coverage. When it launched its second-generation blade battery and flash-charging technology on March 5, 2026, it had built 4,239 flash-charging stations. Just over five months later, that number rose to 10,000. In July and August alone, Byd Company Limited (002594.SZ) added roughly 3,000 stations.

Of the 20,000-station plan, 18,000 will be "station-within-station" sites and 2,000 will be highway flash-charging stations. By end of August, the highway stations had neared the 2,000 target, with the 2,000th site about to come online.

Byd Company Limited (002594.SZ)'s first batch of stations was about building a national skeleton. It organized an effort dubbed "one center, seven war zones, hundred-regiment campaign," going from the first station to 4,239 in just three months. The flash-charging piles, hosts, energy storage, and control software are developed and manufactured in-house, allowing equipment manufacturing, construction design, and vehicle launch timing to be pushed forward within a single organization.

Its own vehicle lineup provided the first source of traffic. Highway, scenic area, and high-altitude stations are hard to recoup costs from local charging volume alone, yet Byd Company Limited (002594.SZ) still moved first to fill these network gaps so that flash-charging models could travel across cities. Once the network opened up, vehicles from other brands began arriving as well.

These network gaps are also the scenarios hardest to bypass as new energy vehicles continue to replace fuel cars. One reason Byd Company Limited (002594.SZ) rushed to build a station in Harbin, where temperatures hit minus 30 degrees Celsius, was to bring the goal of "charging as fast as refueling" into extreme low-temperature settings, directly addressing consumer concerns about winter charging.

According to Byd Company Limited (002594.SZ), after the flash-charging technology launch, station construction demand quickly crowded into urban areas. It rolled out a program called "Dream Station Building," where four vehicle owners applying simultaneously with a suitable site could push a station forward. It received tens of thousands of applications in March alone. By the time the 10,000th station was completed, 4,228 of them had come through this program. A large share of applications landed in core urban districts where power capacity and land are both tighter, slowing the previous construction pace.

Byd Company Limited (002594.SZ) subsequently leaned more on existing social stations, entering urban areas through the "station-within-station" model, with partnerships involving Sinopec, PetroChina, Shell, and TELD. These sites reuse the original land, power distribution, and daily operations, while Byd Company Limited (002594.SZ) brings in flash-charging equipment and its own users, allowing it to leverage existing locations and partial power conditions. Of the 20,000-station plan, 18,000 adopt this approach, with the focus of new additions now firmly on urban densification.

Times EV Service does not own vehicles, so it first needs automakers to adapt models for battery swapping. Swapping is not just about interface compatibility; the vehicle's chassis, battery dimensions, and thermal management all need to be adjusted in advance. If the network is too sparse, automakers will not readily write swapping into their product definitions; without enough compatible models, the stations built first will lack stable traffic.

Times EV Service's strategy is "stations before vehicles" — first lay out the basic network, then push partner models in. Early stations were sited based on urban heat mapping, with layout accelerated according to automaker sales feedback. Sales personnel at relevant automakers note that many buyers are coming specifically for the swap capability.

Times EV Service's end goal is to get users to a station within three minutes in high-density areas. With sufficiently dense stations within cities, swapping can enter dealership sales conversations, and automakers introducing the next compatible model will have an existing network ready to use.

Now CATL (03750.HK) is starting to connect these urban nodes. Its plans this year include covering nearly 190 cities with 4,000 "super-swap integrated stations" and laying out a "12 vertical, 11 horizontal" highway network.

Sources close to CATL (03750.HK) reveal that the next focus for swap station deployment will be highways, with relevant projects landing soon. Compared to urban stations, highway sites face service area location and approval constraints, making construction slower.

Byd Company Limited (002594.SZ) already has a huge and widely distributed base of owned vehicles, so highway stations can directly absorb cross-city demand once built. Times EV Service's highway stations, by contrast, must connect cities that already have density at both ends, and they also need vehicles that can use the same standard battery. Without urban stations and compatible models first taking shape, highway routes will struggle to form continuous traffic flow.

NIO put swap stations on highways earlier, while Li Auto and XPeng have expanded their super-charging networks alongside their high-voltage models. In this round, Byd Company Limited (002594.SZ) is shifting from a national skeleton to urban densification, while Times EV Service is moving from urban density to highway connectivity.

The Cost of Building the Network First

Charging networks face an unavoidable time lag: stations must arrive before vehicles, but revenue only appears once vehicles enter the station. The very scenarios that most effectively ease fuel car user concerns — highways, extreme cold, and remote areas — often require the heaviest upfront investment.

Build too slowly, and users will not treat flash charging or swapping as a reliable option, nor will automakers redesign models for a sparse network; build too fast, and equipment, site leases, and battery depreciation hit first, with traffic possibly taking years to catch up. Every brand building its own network bears this upfront cost, regardless of the technology route chosen.

From Byd Company Limited (002594.SZ), building stations in frigid regions adds construction difficulty. To meet winter deadlines, its crews used insulation tents and electric heating blankets to compress a planned 50-day schedule down to 28 days; at one site in Hefei where geological conditions prevented further excavation, prefabricated steel frames cut construction time for similar sites from several weeks to a few days. Climate and geological issues can be engineered into replicable construction playbooks for the next station.

Core urban districts are much harder to replicate. After the flood of applications in March, Byd Company Limited (002594.SZ) expected construction to accelerate further, but the pace actually slowed. Equipment can be manufactured in bulk, but urban power capacity, underground pipelines, merchants, property management, and community relations must all be coordinated site by site. The journey from 4,239 to 10,000 flash-charging stations saw the bottleneck shift from building equipment to securing a truly usable space.

The public charging industry is already showing the strain of this math. Equipment, power distribution, and site leases all require upfront capital. China Charging Alliance monitoring data shows average public pile utilization at around 6.2% in Q4 2025. As of late July this year, 18.584 million private charging guns nationwide were absorbing a large share of daily charging, meaning public stations must compete for time-sensitive traffic from long-distance drivers, ride-hailing vehicles, and logistics fleets.

Those orders are also concentrated in a narrow window of hours. Holiday highway travel demands high-power equipment to digest short bursts of traffic, while the same equipment may sit idle for most of a workday; urban stations rarely run at full capacity around the clock. To keep users willing to rely on the network even during congestion, operators must build in redundancy ahead of time.

Plateau, Gobi, desert, and some scenic-route stations can complete the network, but their utilization rarely rivals urban sites. One charging pile company executive notes these locations can hardly recoup costs from local charging volume alone. They serve vehicle sales and the broader network, yet individually their books may remain inefficient for long stretches.

Battery swapping carries even heavier assets. Beyond site and equipment, stations must stock rotating batteries; without enough compatible vehicles, batteries depreciate in the warehouse, and too many standards can fragment storage capacity.

CATL (03750.HK)'s 2025 annual report states that its Chocolate swap network has already achieved profitability in Chongqing.

However, whether users dare to depend on one charging method depends on whether stations exist where they frequently go, and also on whether gaps appear during occasional cross-city trips. Edge stations, even with low traffic, may be an indispensable link in the whole network.

Times EV Service therefore emphasizes accounting at the city and network level. Automakers like Byd Company Limited (002594.SZ), NIO, and Li Auto also fold some station network investment into their vehicle sales and user service ledgers.

Station density produces two opposing effects at once: an additional station in the same area may siphon orders from neighboring sites, but it also shortens the time users spend searching and waiting, which increases vehicle sales and overall network usage. If only single-station profit is measured, companies will under-build stations that preserve network integrity; if coverage is the only goal, capital gets sunk into locations that never see traffic.

What companies find even harder to calculate is whether the incremental vehicle sales, user retention, and standard expansion generated by new stations can offset the dilution of utilization at surrounding sites.

This raises the barrier to entry for charging networks even further. Ordinary operators rely mainly on electricity and service fees to recover investment; automakers can earn returns from vehicle sales and user retention; battery makers can also benefit from installation, leasing, and standard expansion.

The latter two players can tolerate longer incubation periods, yet they cannot escape utilization either. Once station scale rises from thousands to tens of thousands, every piece of idle equipment, idle battery, and overestimated traffic flow gets magnified simultaneously.

The Watershed Moment for Station Networks

New stations are hard to monetize in the short term, yet the top players are accelerating this year, signaling that companies are now accounting for station network investment together with vehicle sales and battery installation. Once a network crosses several thousand or tens of thousands of stations, the assessment shifts beyond whether stations exist to who can bring stable traffic and who can connect vehicle models and orders.

Even as driving range keeps growing, charging convenience still influences whether users choose pure electric, plug-in hybrid, or extended-range vehicles.

That shift is already visible at the retail level. A 4S store investor tells us that before a certain automaker's swap version launched, he remained uncertain about the car's sales prospects. Yet within two months of launch, the swap version accounted for the overwhelming majority of sales. The dense deployment of swap networks eased consumers' "range anxiety" about swap-enabled models.

NIO completed its 100 millionth battery swap in February, and Li Auto and XPeng's self-operated super-charging networks have both reached the thousands-of-stations scale. Self-built charging has evolved from a vehicle accessory into a long-term operating business.

Roland Berger's "EV Charging Index 2026," released in late July 2026, shows global new public charging points grew by about 1.1 million in 2025, slightly fewer than in prior years, as mature markets shift focus from network expansion toward utilization, fast-charging quality, and commercial sustainability. China is still accelerating station construction, but the larger the network, the harder it is for station counts to mask the realities of utilization, equipment availability, and payback cycles.

Users' actual choices also make it difficult for any charging network to remain closed for long. The "2025 EV User Charging Behavior Study" released by the China Charging Alliance in July this year shows 95.4% of users prefer DC fast charging as their first choice, 66.85% favor stations with supporting services, and 87% charge across multiple operators, using an average of six. Users follow location, power, price, and service — they do not stay locked into one brand ecosystem.

That fluidity has already appeared at Byd Company Limited (002594.SZ)'s new stations. The company discloses that as of end of August, users from other brands accounted for nearly one-third of flash-charging station traffic; since the "Flash Charge China" strategy launch, the open network has accumulated over 210 million kWh of charging volume in less than six months of operation.

Users may cross networks, but first-mover advantages have not disappeared. Orders can flow across brands; core locations and connected power capacity do not. Of Byd Company Limited (002594.SZ)'s 20,000-station plan, 18,000 are "station-within-station" sites, meaning most of the new network will sit inside existing social stations.

In these partnerships, Byd Company Limited (002594.SZ) brings its own models, vehicle-machine interfaces, and charging entitlements, while operators gain incremental traffic, with some order entry points flowing into the vehicle system and brand app.

Battery swapping pushes the first-mover advantage even further upstream into vehicle R&D. At its Super Tech Day in April, CATL (03750.HK) disclosed that the Chocolate swap standard is collaborating with 11 automakers and 18 passenger car brands across 25 models, forming the base for a unified standard. Every additional compatible model that launches may simultaneously add power battery orders and future swap users; the more models and vehicles using the same battery specification, the easier it is to rotate station batteries, and new automakers joining do not have to bear the full cost of a network alone.

The Ministry of Industry and Information Technology's "2026 Automotive Standardization Work Points" continues to advance review and approval of chassis swap standards and swap compatibility research, meaning industry-wide common rules are not yet settled. CATL (03750.HK)'s proposed 100,000 shared charging and swapping facilities will rely on daily traffic from these models. Which swap standard is chosen at the model development stage extends the battery maker's relationship with automakers from the installation stage into post-delivery charging services.

In the past, automakers could catch up on technology gaps within one product generation, and market rankings could reshuffle after a new model launch. Charging networks have no such rhythm. There must first be enough vehicles on the road for stations to generate stable orders and operators to dare to keep densifying; as stations densify, purchase hesitation among the next wave of users drops. Once this cycle takes hold, the cars sold in the previous round are paving the way for the next round of new vehicles.

A blockbuster model can quickly rewrite sales figures, but it is very difficult to fill in core locations, power distribution capacity, and vehicle model compatibility within a single product cycle.