CATL can reach 3,000 swap stations without Huawei. The cars are the constraint

CATL added more than 520 stations in 91 days and is on pace for 3,000 this year. Its Huawei framework therefore needs compatible car models, not more stations.

CATL does not need Huawei to build its battery swap stations. It needs cars that can use them. On Monday it signed a framework with Yinwang to work on that, but the agreement names no car model.

Yinwang Intelligent Technology is the Huawei affiliated smart car business. The third signer was Shidai Dianfu, CATL's own battery swapping unit. The framework matters more to CATL than to Yinwang, because cars are the scarce input.

CATL needs customers because its shares fell to a one year low in September. That is about 35% below their early May high. Investors fear that carmakers are moving cell orders, meaning orders for the battery units inside a pack, to other suppliers.

Yinwang brings Huawei branded vehicles at the moment CATL's share price shows customers drifting toward other cell makers. For Yinwang, the benefit is a second energy answer beside Huawei's own supercharging.

Swapping adds to charging and does not replace it. Huawei's announcement says Qiankun intelligent driving, its driver assistance system, already connects to a charging ecosystem it puts at a million piles. The parties pledge equal and mutually beneficial cooperation with shared resources. That language commits no capital and sets no date.

What was signed is narrow. According to the framework text as published on Monday, the three parties agreed to cooperate in two directions.

The first direction links the vehicle, the station and the cloud. It does this by joining CATL's Chocolate Battery Swap network to Huawei's Qiankun intelligent driving system. The second has Yinwang formally join the swap ecosystem and work with partners on adapting cooperating models to the network.

Huawei's Qiankun intelligent car solution unit announced the agreement on Monday. Its title was Intelligent Energy Replenishment Ecosystem Strategic Cooperation Agreement. The signing ceremony took place in Ningde. CATL chairman Zeng Yuqun, Yinwang chief executive Jin Yuzhi and CATL co president Yang Jun attended.

The parties describe a car that finds the station, parks and completes a 99 second automatic swap on its own.

The framework does not mark Huawei's entry into battery swapping. CATL had 1,650 stations at the end of May and 2,000 on 30 June, about 11 a day across June.

By 29 September the count stood above 2,520. That is a pace of at least 5.7 a day over 91 days. Reaching the 3,000 promised for the end of the year takes 480 more stations in 93 days, or 5.2 a day.

The third quarter build was slower than June, which carried some catch up. It still ran ahead of what the target requires. At 5.7 a day, the network would pass 3,000 with about 50 stations to spare.

The 2,520 figure is a floor, so the true count is higher. CATL reached nearly 1,000 new stations in the first half of 2026, roughly as many as in all of 2025. So the plan did not depend on a partner paying for more.

The network already spans 31 provinces and about 200 cities. A new entrant would take years to match that footprint. The target itself, more than 3,000, is the same one given in both reports.

The missing input is cars. The Paper reported that 2025 brought 10 new models from five automakers: FAW, Changan, BAIC, Chery and GAC. In April CATL announced cooperation with 11 automakers and 18 passenger car brands on 25 models. That is a respectable start for a young standard.

But a station earns its keep only when enough compatible cars pass through it. Yinwang's partner brands would add vehicles that sell on Huawei's software. That is a different customer pool from the carmakers already signed, and the framework names none of them.

Dongfeng and Huawei's Epicland brand is reportedly weighing a swap version of its second model. But CnEVPost reported that the work is at a pre research stage with no decision made.

CATL needs that pool for a reason beyond keeping stations busy. In September, CnEVPost also reported that Li Auto had turned to Sunwoda cells, while Xiaomi and Aito had added CALB and Gotion. It said this diversification was behind the share price slide.

CATL still supplied 41.45% of installed capacity in August. Installed capacity is the battery capacity fitted to new cars. So the fear concerns direction rather than current position.

A swap network changes the arithmetic of that direction. A car built to take a swappable pack in CATL's format is tied to CATL's cells. An owner who relies on the stations cannot change supplier without giving up the network.

That lock in is also the price an automaker pays. A Yinwang partner brand that designs its next platform for a swappable pack gains access to a network of about 2,500 stations. It also gets the 99 second swap. But it gives up leverage over its cell supplier.

A fixed pack with fast charging keeps that leverage, but it depends on how quickly chargers spread. For a 2027 model, the choice has to be made at platform freeze, the point when the car's basic design is locked. That is within months, so the framework puts the question to Yinwang's partners sooner than it answers it.

CATL has turned a network race it was already winning into a vehicle race it has not yet won. Stations can be counted every quarter. Compatible cars arrive only when an automaker commits a platform, and that commitment is what Monday's signing is meant to start.

CATL should reach 3,000 stations on schedule. The network is likely to be at or past that mark when the year closes.

The stronger signal will come by the end of the first quarter of 2027. A Yinwang badged model should either appear at a swap station by then or not. If none appears, the framework will have been a statement about CATL's customer list rather than Huawei's energy strategy.