Where Chinese EV makers are actually building now, which Western plants they are buying, and how fast their share of the car market is moving outside their own.
Tariffs are the easy story. The harder one is the map of Chinese EV production outside China. In 2026, BYD, Geely, Great Wall, Chery and SAIC are not shipping finished cars from Shenzhen into contested markets; they are opening plants in Thailand, Brazil, Hungary, Indonesia, Spain and Russia, often on the exact sites where Ford, Daimler, GM, Renault and Nissan used to build. The move is the story.
BYD is the most visible, but it is not alone. Of the roughly ten overseas plants already running, BYD holds four (Rayong in Thailand, Camaçari in Brazil, Szeged in Hungary, Subang in Indonesia), Great Wall three (Tula in Russia, Rayong in Thailand, Iracemápolis in Brazil), Chery one (the Ebro joint venture in Barcelona, Spain), and Geely runs its Volvo assets in Sweden, Belgium and South Carolina. SAIC's first European plant is under construction in Galicia for a 2028 start. The pattern is not scattergun. Every one of these sites sits inside or next to a market where the Chinese brand already sells, and every one carries assembly capacity in the 50,000 to 200,000-vehicles-a-year range: large enough to serve a region, small enough to keep converting as demand clears tariffs.
BYD's Camaçari plant in Bahia is the largest Chinese automotive investment outside Asia, with a nameplate capacity of 150,000 electric and hybrid vehicles a year, scalable to 300,000. The Hungarian plant at Szeged, in trial production since January 2026 and scheduled for full series output in Q2 2026, lands at around 200,000 vehicles a year and is the company's first "Made in Europe" line.
Everyone else is converting. Great Wall's Iracemápolis plant was a Daimler site before 2021. Chery's Spanish partner Ebro took the former Nissan plant in Barcelona and converted it on a 150-million-euro bill, 60% Ebro, 40% Chery. Great Wall's Rayong line is a former GM plant. Geely has gone further and bought a 26.4% stake in Renault Brasil, part of a R$3.8 billion joint investment at the Ayrton Senna complex in São José dos Pinhais. The pattern is efficient: the car is new, the paint shop and the stamping line are not.
Europe is the slower frontier: Chinese brands crossed the 4% market share line in March 2026 (the exact number varies across sources, mostly because the definition of "European market" shifts and because MG, Polestar and Volvo are sometimes read as Chinese and sometimes not), and the planned Spanish, Hungarian and Turkish plants are the response. Southeast Asia is further along, with Chinese OEMs reaching about 15.7% of the ASEAN-6 market in the first half of 2026, which explains the two Thai plants and the two Indonesian ones. Brazil is where the story is loudest: electrified vehicles climbed to 25.2% of the car market in 2026, a share change European policy debate still tends to project forward rather than report back. The US sits outside this map; tariffs on Chinese finished vehicles have kept the share effectively at zero, which is why no Chinese OEM has announced a US assembly plant and Geely's Ridgeville line continues to be its only US-flagged capacity.
Interactive tool
Fifteen plants: ten operating, four planned or under build, one on hold. Filter by company, region or status. Click a card for the primary source.
No plants match that combination.
A regional EV plant is a bet on being able to source cathodes, anodes, magnets and power electronics within a reasonable supply radius. On the current data, that bet still points back to China for most of the stages that matter. Our Mineral Flows Map shows roughly 95% of LFP cathode manufacturing sitting inside China (IEA, Benchmark), the chemistry the Camaçari and Rayong lines mostly build on, and the chain-diagram view for nickel ends in Chinese cathode manufacturing even when the ore is Indonesian or Brazilian. A Hungarian plant with LFP cells does not escape that geometry; neither does a Thai one. The companion research paper walks through why.
That is the point industrial-policy conversations in Brussels, Brasília and Jakarta have begun to focus on. A plant on your soil is one thing. The supply pipe behind it is another. The Hungarian and Spanish plants will reopen the European argument about whether building the cell plant is enough, or whether the cathode-active-material step needs to come too; the Brazilian and Indonesian plants will reopen the same argument on the ore end, which is where Africa's and South America's reserves now re-enter the frame.
First, no claim about profitability, and no claim that every plant is running flat out. BYD Rayong is reported to be running at around a third of nameplate as it passes its first 100,000 units, which is a story about demand ramp and local-content rules, not a story about Chinese OEMs being unable to make money in Thailand. The newer Brazilian, Hungarian and Indonesian lines are in their first year of production and should not be read against mature-plant utilisation. We report capacity; we do not infer utilisation from it.
Second, no claim about job creation net of displacement. The Camaçari and Iracemápolis sites were idle or closing before the Chinese acquisitions, so a job gained at a reopened plant is not a like-for-like replacement of a job lost at the same site. Independent impact tracking of the kind the Lab builds for European funders would be the right instrument here.
Third, no projection. Manisa in Turkey is listed as on hold; Lembah Beringin in Malaysia will finish in 2027; the Galicia start is slated for 2028. These are announcements, not deliveries. We report the planning status and leave the forward curve to the companies to defend.
Plant capacity, site history and start dates are drawn from company announcements and 2024-2026 trade press. The regional market-share figures come from the three industry associations most reliable at this cadence: the European Automobile Manufacturers' Association (ACEA) for Europe, ANFAVEA for Brazil, and the ASEAN Automotive Federation (AAF) for Southeast Asia. The LFP and cathode-concentration figures are the Lab's own Mineral Flows Map, drawn from the IEA Global Critical Minerals Outlook 2026 and Benchmark Mineral Intelligence tracker series. Every claim with a number is anchored to a bibliography entry below. Entries 1–13 are plant-level sources; 14–16 are the share statistics; 17–18 are the mineral-chain references.
For the full bibliography across every study behind the chain, see the companion research paper.
Editor's note. The exact text of the research notes behind this piece, pasted verbatim and unedited for transparency. Downloadable as PDF or DOCX. This block is temporary and will be removed once the article has settled.
Existing Factories of Chinese automotive companies abroad:
BYD:
Rayong, Thailand: production began in July 2024, with a designated annual capacity of 150,000 vehicles. Actual use is 30 per cent of designated capacity.
Camaçari, Brazil (a former Ford site): the Camaçari complex began assembling vehicles in July 2025. The annual capacity of operations is 150,000 electric and hybrid cars a year, which can be scaled up to double the amount. It is BYD's largest investment outside of Asia.
Szeged, Hungary: trial for passenger vehicle production started in January 2026, with full series production scheduled for Q2 2026 and an eventual capacity of about 200,000 vehicles a year. The company's first "Made in Europe" production.
Subang, Indonesia: BYD opened a second regional plant on September 3 with a capacity of 150,000 vehicles a year.
Great Wall Motor (GWM):
Tula, Russia: 500 million dollar investment with an annual capacity of 150,000 vehicles, producing since 2019.
Rayong, Thailand: annual capacity of 80,000 vehicles opened in June 2021 on a former GM site.
Iracemápolis, Brazil: converted from a Daimler plant acquired in 2021, with an annual capacity of 50,000 vehicles. Production started in August 2025.
Chery:
Barcelona, Spain (Ebro joint venture, former Nissan plant): Ebro holds 60% and Chery 40%. The partners have invested more than 150 million Euro and the plant could produce up to 200,000 vehicles annually.
Geely (through Volvo Cars):
Volvo is 78.7% owned by Geely Holding and runs plants in Torslanda, Sweden and Ghent, Belgium. The plant in Ridgeville, South Carolina (US) has installed capacity of 150,000 vehicles a year.
Planned investments of Chinese automotive companies abroad:
BYD:
Manisa, Turkey: construction of a plant with an investment of 1 billion dollar and annual capacity of 150,000 vehicles is on hold.
SAIC (MG):
Galicia, Spain: first factory in Europe, with a 200 million Euro investment, annual capacity of 120,000 vehicles. Start of operation in 2028.
Chery:
Lembah Beringin, Malaysia: industrial park to finish in 2027.
Barcelona, Spain: Chery wants to add capacity at its Barcelona joint venture.
Geely:
São José dos Pinhais, Brazil (Renault plant): Geely bought a 26,4% stake in Renault Brasil. The partners announced a joint investment of 3.8 billion R (real) at the Ayrton Senna complex to produce new car models.
Market shares of Chinese automotive companies across continents:
Europe: as of March 2026, Chinese automotive companies are set to cross the 4% market share barrier in Europe.
US: no percentages found in reports. Chinese manufacturers are largely shut out by high tariffs on Chinese imports.
South America (Brazil): the market share for electrified vehicles climbed to 25.2%.
Southeast Asia: Chinese OEMs reached about 15.7% of the ASEAN-6 market in H1 2026.
Sources and references
Every claim above is backed by one of the numbered sources below. Shares vary across sources depending on how "Europe" is defined and which brands count as Chinese.
If this reads close to a decision you are weighing, we would like to hear about it. A short note is enough to start.
The Lab is an independent research team, not an advocacy organisation. We do not campaign, endorse products, or accept commissions where a finding is expected in advance. Where a partner's own theory of change is what needs testing, that is a legitimate engagement, and the Lab tests it honestly.