XPeng's 60,000-car month: what has to go right

The gap is nearly 19,000 cars

XPeng finished September with 41,256 deliveries. A 60,000 month would take another 18,744 cars, about 45% more, and go well past its busiest month on record, October 2025's 42,013 (XPulse's arithmetic on the September results).

On the 24 August earnings call, He Xiaopeng set the challenge of passing 60,000 sales in a single month during the fourth quarter. It is a challenge, not formal guidance, which comes with the third-quarter results in November. XPulse measures it against XPeng's monthly deliveries.

The L03 offers the clearest route

The MONA L03 delivered more than 14,000 cars in September, its second full month on sale, XPeng said on 2 October. XPeng has pointed to supply, not demand, as the constraint. On 7 August Jiang Wen, head of research and development for the MONA series, said chip supply had disrupted August's production but that the problems were gradually being overcome. The line had moved to two shifts, and he expected production to reach full capacity in September and October.

Two shifts were already running by early August, so September's figure already reflects the second shift; it cannot be counted again. XPeng has given no monthly L03 production limit.

What would the factories need to produce?

The factory research gives a way to test the scale. XPeng builds cars at Zhaoqing and in Guangzhou, which has a North and a South Area; the South Area reached mass production in May 2025. Wuhan, XPeng's components base, makes drive motors and battery packs and, Hubei Daily reported in March, also produces and exports components for XPeng cars assembled abroad from kits. It supports production both in China and overseas, but it does not finish cars itself, so its output cannot simply be added to the car counts below.

As a starting point: if every car the passenger-car association CPCA counted as exported in 2025 was also built that year, XPeng's plants built about 449,000 cars in 2025, an average of roughly 37,400 a month. That is XPulse's calculation and a guide to the past, not a measure of today's capacity.

XPulse illustration: take a busy month with 25 production days. Assume each line keeps its stated or designed pace for the hours below, with enough parts and no output lost to stops or rework. This is a demanding best case, not XPeng's disclosed fourth-quarter capacity.

Zhaoqing: 20,000 cars. XPeng has described a pace of one car every 90 seconds, or 40 an hour. At 20 productive hours a day, which is XPulse's assumption, 25 days give 20,000. There is a precedent: the MONA M03 reached a production rhythm of 20,000 a month before the 2025 Lunar New Year.

Guangzhou North: 12,000 cars. Its final-assembly line is designed for 30 an hour, and two eight-hour shifts over 25 days give 12,000. A third shift is permitted, but its approval also adds unfinished car bodies, so the extra hours cannot all be counted as cars.

Guangzhou South: 28,000 would close the gap. That means 56 finished cars an hour over 20 productive hours and 25 days. Its original assembly design is 60 an hour, or 30,000 on the same assumptions, but its permits chiefly describe bodies and components, and nothing published establishes a current finished-car rate of 60 an hour.

Zhaoqing and Guangzhou North would contribute 32,000 between them, leaving 28,000 for the South Area. At the 30,000 its design allows, the total would be 62,000, a margin of just 2,000 cars, and only if that pace turns into finished cars. That makes the South Area the deciding unknown.

Partners in Austria, Indonesia and Malaysia also assemble XPeng cars, and those count in deliveries too, so the Chinese plants need not build all 60,000. XPeng does not say how many are assembled abroad.

South's expansion could give Guangzhou more room

The South Area's next stage would double its planned output of car bodies, from 300,000 to 600,000 a year. It also adds assembly equipment for another 60 cars an hour, paint lines to match, and a line making its own battery packs, 300,000 a year. A car can be finished only as fast as its body is painted and its battery arrives, so if some of the new paint or battery lines run reliably in the fourth quarter, the existing assembly line could work closer to its full pace.

No start-of-production date has been published: a 12-month construction plan approved in December 2025 does not establish completion by December 2026, and car bodies are not finished cars. The Guangzhou guide sets out what is operating.

The new models have to add to the total

The G9L began deliveries on 22 September, so October is its first full month. The MONA L05, listed as pending at Guangzhou, is due to launch in China in the fourth quarter; what it adds depends on when deliveries begin.

September's 41,256 already included 7,224 GX, more than 14,000 L03s and the cars sold abroad. Gains have to come on top of those and outweigh any decline in the M03 or other models.

Orders are strong; the home market is not

On 24 August, with the quarter not yet over, XPeng said new firm orders in the third quarter were 50% above the second, a record. On 30 August an XPeng vice-president said daily L03 orders were still far above deliveries.

The market around it is shrinking. CPCA counted China's retail sales of new-energy cars in 1–27 September 20% below the same days of 2025. The M03, about 41% of XPeng's deliveries in 2025, has also lost ground since the L03 arrived: Deutsche Bank estimated that new M03 orders fell about 40% after the L03's launch. The figures do not show how much of that is buyers choosing the L03 rather than weaker demand, but an L03 that replaces an M03 adds nothing to the total.

XPeng is aiming for more than 40,000 abroad

XPeng says overseas deliveries passed 10,000 in September for the first time, without giving the exact figure. In August it said the L03's overseas ramp in the fourth quarter would lift quarterly overseas sales above 40,000, against more than 20,000 in the second quarter. L03 deliveries abroad have already started in several markets.

XPulse arithmetic: more than 40,000 in the fourth quarter means an average above 13,300 a month. The months need not be equal, and a strong December abroad could help the worldwide total past 60,000.

Cars sold abroad are counted in the same 60,000, not on top of it, so the 13,300 a month should not be added to the factory figures above. Growth abroad helps by finding buyers for more of the cars XPeng builds.

Timing matters too. A car shipped abroad in December may reach its buyer only in January. But cars built in October and November can be delivered in December, so a 60,000 delivery month does not need 60,000 cars built in that month.

October will show whether the route is opening

October's deliveries, due on 1 November, are the first test. A higher total with more L03s would support the supply story, and the G9L's first full month will show what it adds. November should bring formal fourth-quarter guidance, and the L05's launch date will show how much of the quarter it has.

Neither October nor November has to reach 60,000 for a December crossing to remain possible, but the later the surge, the more has to happen in less time.

XPulse's reading: the route to 60,000 runs through fuller, steadier factory output, above all in Guangzhou's South Area, a stronger L03 ramp and new models that add to the total rather than replace older ones, with growth abroad helping. The next two monthly reports will show whether that is turning into deliveries.

Learn more about XPeng's factories