Chinese EV Companies 2026
By Axis Intelligence Research
Co-author: Aidan Jad, EV & Clean Energy | Last updated: September 1, 2026 | License: CC BY 4.0
Six listed Chinese EV makers delivered 4,138,570 vehicles in the first half of 2026 and between them earned RMB 13.29 billion. Axis Intelligence Research measures a Profit Per Vehicle spread of 11.6 times between the best and worst performer that still made money — and a gap of RMB 27,457 per car between BYD and Li Auto.
Quick Answer
BYD remains China’s largest EV company, with 1,808,511 new energy vehicles sold in the first half of 2026 and net profit attributable to shareholders of RMB 12.33 billion. Geely follows at 1.42 million vehicles. Axis Intelligence Research calculates a Profit Per Vehicle (PPV) of RMB 6,815 for BYD, RMB 589 for Leapmotor, and negative RMB 20,642 for Li Auto — the widest per-unit dispersion the sector has recorded in a reporting period.
Key Findings
- Axis Intelligence Research calculates a Profit Per Vehicle (PPV) of RMB 6,815 for BYD in the first half of 2026, down 5.71% from RMB 7,228 a year earlier — the first decline in BYD’s per-unit profitability at a time when its volumes also fell 15.72%.
- Axis Intelligence Research finds that Leapmotor, the profitable Chinese EV startup, earned RMB 589 per vehicle — about US$87 per car, or 0.55% of revenue. Profitability and profit are not the same achievement.
- Axis Intelligence Research calculates a Unit Breakeven Gap (UBG) of 166,377 vehicles for Li Auto — at its first-half gross profit of RMB 24,004 per car, Li Auto would have needed to sell 86.0% more vehicles than it did to erase its net loss.
- According to filings with the U.S. Securities and Exchange Commission, Li Auto, NIO and XPeng together delivered 550,572 vehicles in the first half of 2026 and recorded a combined net loss attributable to shareholders of RMB 8.33 billion.
- According to NIO’s second-quarter filing, the company narrowed its loss per vehicle by 94.0% year on year, from a first-half 2025 reading of RMB 105,408 to RMB 6,371 — the largest single-year improvement in unit economics in the group.
Which Chinese EV Companies Sold the Most Vehicles in the First Half of 2026?
Two companies account for more than three-quarters of the volume in this dataset, and the ranking has not changed at the top. What has changed is that the leaders are now shrinking while the challengers grow.
| Rank | Company | H1 2026 vehicles | Change vs H1 2025 | Scope | Source |
|---|---|---|---|---|---|
| 1 | BYD | 1,808,511 | −15.72% | New energy vehicles, wholesale | BYD (HKEX filing) |
| 2 | Geely Automobile | 1,423,000 | +1% | All vehicles, including combustion | Geely Automobile Holdings |
| 3 | Leapmotor | 356,487 | +60.8% | All vehicles, BEV and EREV | Zhejiang Leapmotor Technology |
| 4 | Li Auto | 193,472 | −5.13% | All vehicles, EREV and BEV | Li Auto (SEC filing) |
| 5 | NIO | 191,123 | +67.43% | NIO, ONVO and FIREFLY | NIO (SEC filing) |
| 6 | Xiaomi | 185,055 | n/a | Smart EV segment | Xiaomi Corporation |
| 7 | XPeng | 165,977 | −15.83% | All vehicles, BEV and EREV | XPeng (SEC filing) |
Three of the seven shipped fewer cars than a year earlier. That includes the market leader. BYD’s 1,808,511 new energy vehicles for the half, disclosed in its voluntary monthly sales announcement to the Hong Kong Stock Exchange, were 337,443 units below the same period of 2025.
The two companies that grew fastest, Leapmotor at 60.8% and NIO at 67.43%, grew from very different bases and with very different price points. Leapmotor’s revenue works out to about RMB 106,900 per vehicle. NIO’s brand ranked first in China’s passenger vehicle market above RMB 350,000 in the second quarter, on the company’s own account.
Note the scope column, because it does most of the analytical work on this page. Geely’s 1.42 million includes combustion vehicles; roughly 799,000 were new energy on the company’s reporting. BYD’s figure is new energy only, because BYD stopped building pure combustion cars in 2022. Comparing the two totals as though they measure the same thing is the single most common error in coverage of this sector.
Mia Scarlett: Volume rankings are the part of this story that gets written, and they are the part that tells you least. Two of the top three by units are shrinking, and one of the fastest growers earns less per car than a mid-market dealer makes on a set of floor mats. The filings say something the delivery press releases do not: scale in this industry stopped guaranteeing anything around 2025.
Source: Axis Intelligence Research — CC BY 4.0
How Much Profit Does Each Chinese EV Company Make Per Vehicle?
This is the question the sector’s own disclosures answer and nobody assembles. So Axis Intelligence Research built the metric.
PPV (Profit Per Vehicle) = net profit attributable to shareholders ÷ vehicles delivered, taken from the same disclosure, for the same period.
Both legs come from one filing. No blending of a company-reported delivery number with a third-party financial estimate, no annualising, no currency conversion inside the calculation. The result is a single figure that can be recomputed by anyone holding the same two numbers.
| Company | Net profit attributable, H1 2026 | Vehicles, H1 2026 | PPV (RMB) | PPV (US$) | Basis |
|---|---|---|---|---|---|
| BYD | RMB 12,325m | 1,808,511 | 6,815 | 1,004 | Net profit attributable, NEV only |
| Geely Automobile | RMB 9,680m (core) | 1,423,000 | 6,803 | 1,003 | Core profit, all vehicles incl. ICE |
| Geely Automobile | RMB 9,090m (net) | 1,423,000 | 6,388 | 941 | Net profit attributable, all vehicles |
| Leapmotor | RMB 210m | 356,487 | 589 | 87 | Net profit attributable |
| NIO | RMB −1,217.6m | 191,123 | −6,371 | −939 | Net loss attributable |
| XPeng | RMB −3,121.2m | 165,977 | −18,805 | −2,771 | Net loss attributable |
| Li Auto | RMB −3,993.7m | 193,472 | −20,642 | −3,042 | Net loss attributable |
Sources: BYD 2026 Interim Report; Geely Automobile Holdings H1 2026 results; Leapmotor 2026 interim results; NIO, XPeng and Li Auto Q2 2026 filings. PPV calculated by Axis Intelligence Research. USD at RMB 6.7851 to US$1.00, the Federal Reserve H.10 rate for June 30, 2026 used by NIO and Li Auto in their own filings.
The full span runs from positive RMB 6,815 to negative RMB 20,642 — a distance of RMB 27,457 for every car that leaves a factory. Two Chinese manufacturers can build broadly similar mid-size electric SUVs, sell them into the same price war, and land RMB 27,457 apart on the bottom line per unit.
What PPV shows that market share does not
Blend the six companies with a disclosed net profit line and Axis Intelligence Research calculates a sector PPV of RMB 3,212 across 4,138,570 vehicles. That blended figure is almost entirely BYD and Geely: the two of them are 78.08% of the volume and effectively all of the profit.
Strip them out and the picture inverts. Axis Intelligence Research finds that the three U.S.-listed startups — Li Auto, NIO and XPeng — delivered 550,572 vehicles in the half at a combined PPV of negative RMB 15,134. Half a million cars, sold at a loss on every one.
The BYD-to-Leapmotor ratio is the reading that deserves the most attention: 11.57 times. Both companies are profitable. Both are Chinese. Both sell heavily into export markets. One earns eleven and a half times as much per car as the other, and the gap is structural rather than cyclical — it reflects vertical integration, battery ownership and product mix, none of which resolve in a quarter.
Why Geely’s number is not BYD’s number
Geely publishes its own version of this metric, and the comparison is instructive. The company reported core profit per vehicle of RMB 6,806, up 45% year on year. Our computation from its disclosed core profit and unit base gives RMB 6,803, a rounding difference on the denominator rather than a disagreement.
But “core profit” excludes after-tax foreign exchange gains and losses and impairment on non-financial assets. On net profit attributable — the line a shareholder actually owns — Geely’s PPV is RMB 6,388. The RMB 415 difference is currency, and it is why Geely’s headline net profit fell about 2% to RMB 9.09 billion while its core profit rose 46% to RMB 9.68 billion in the same six months.
Both figures are correct. They answer different questions. We publish both, on separate rows, rather than choosing the flattering one.
A calculation Axis Intelligence Research declines to publish: we do not compute a PPV for Geely on NEV volume alone. Geely’s profit is a group figure covering combustion and electric vehicles together; dividing a whole-company profit by a partial volume would manufacture a per-EV number the company has never disclosed and cannot be held to. The same restraint applies to any manufacturer with a mixed powertrain book.
Aidan Jad: The per-unit number is where the price war stops being an abstraction. Chinese domestic NEV demand fell in the first half while penetration hit records, which means every manufacturer is fighting for a share of a shrinking pool of buyers, and the weapon is price. Twenty thousand yuan of loss per car is what that fight costs a company without a battery plant of its own. BYD makes its own cells. Geely has scale and a combustion business still throwing off cash. Everyone else is buying their most expensive component from someone else at whatever the market says this month.
How Many More Cars Would Li Auto, NIO and XPeng Need to Sell to Break Even?
A loss is a number. The interesting question is what it would take to close it. Axis Intelligence Research built a second metric for that, and it is the one we expect to be quoted most.
UBG (Unit Breakeven Gap) = |net loss attributable| ÷ gross profit per vehicle, same period.
The logic is deliberately simple: hold operating expenses where they are, hold gross profit per car where it is, and ask how many additional units would absorb the loss. It is a sensitivity reading, not a forecast — no company sells 86% more cars without spending more to do it — but it converts an abstract loss into the one unit the industry actually manages by.
| Company | Net loss attributable | Gross profit per vehicle | UBG (units) | UBG as % of H1 deliveries |
|---|---|---|---|---|
| NIO | RMB 1,217.6m | RMB 56,328 | 21,616 | 11.3% |
| XPeng | RMB 3,121.2m | RMB 40,765 | 76,565 | 46.1% |
| Li Auto | RMB 3,993.7m | RMB 24,004 | 166,377 | 86.0% |
Gross profit and net loss from company Q2 2026 filings. UBG calculated by Axis Intelligence Research. Baseline reading.
NIO’s gap is 21,616 vehicles — roughly seven weeks of production at its recent run rate. XPeng’s is 76,565, about four and a half months. Li Auto’s is 166,377, which is to say Li Auto would need to nearly double its half-year volume before the loss disappeared.
The ordering is not the ordering of the losses. Li Auto lost RMB 3.99 billion and XPeng RMB 3.12 billion, a difference of 28%. Their UBGs differ by 117%, because Li Auto’s gross profit per vehicle collapsed to RMB 24,004 while XPeng held RMB 40,765 and NIO delivered RMB 56,328. The denominator, not the loss, is what determines whether a company can grow its way out.
That collapse is visible in the quarterly line. Li Auto’s vehicle margin fell to 9.4% in the second quarter of 2026 from 19.4% a year earlier, roughly halving in twelve months. NIO moved the other way, from 10.3% to 18.5%. Those two trajectories crossed inside a single year.
Mia Scarlett: UBG is the number I would want before reading any of these companies’ guidance. Li Auto guided 95,000 to 100,000 vehicles for the third quarter, which is roughly flat. Nothing in a flat quarter closes an 86% gap. NIO guided 108,000 to 111,000 against a gap worth eleven percent of a half-year — that one is arithmetically reachable, and the company’s non-GAAP line has already crossed zero twice. The distinction between “loss-making” and “loss-making at a distance it can cover” is the whole investment case, and the headline loss figure hides it completely.
Which Chinese EV Startups Are Actually Profitable in 2026?
One, on a statutory basis, and barely.
Leapmotor reported net profit attributable to equity holders of RMB 210 million for the first half of 2026 on revenue of RMB 38.11 billion — a net margin of 0.55%, and its third consecutive profitable half-year. Deliveries rose 60.8% to 356,487 units. Gross margin fell to 11.7% from 14.1%, which the company attributed to raw material costs and product mix.
Run the per-unit arithmetic and the achievement resizes itself. RMB 589 per vehicle. At the June 30 exchange rate that is US$87 a car. It is a real profit, audited and repeated across three halves, and it is also thinner than the margin on a single option package at a German manufacturer.
NIO’s position is the more contested one, and it depends entirely on which line you read. Statutory net loss attributable for the half: RMB 1.22 billion. Adjusted net profit excluding share-based compensation: positive RMB 69.55 million, with the second quarter marking a second consecutive non-GAAP profit at RMB 26.06 million. Share-based compensation of RMB 929.6 million for the half is the entire difference.
We report both and rank on the statutory figure. Share-based compensation is a non-cash charge, but it is a real transfer of ownership from existing shareholders, and a metric that treats it as free would make every high-growth company look better than it is.
| Company | Statutory H1 2026 result | Non-GAAP / core result | Difference explained by |
|---|---|---|---|
| Leapmotor | +RMB 210m | not separately disclosed | — |
| NIO | −RMB 1,217.6m | +RMB 69.6m | Share-based compensation RMB 929.6m |
| Geely Automobile | +RMB 9,090m | +RMB 9,680m core | FX and impairment |
| XPeng | −RMB 3,121.2m | −RMB 2,923.0m | SBC and derivative fair value |
| Li Auto | −RMB 3,993.7m | −RMB 3,618.9m | Share-based compensation |
Sources: company Q2 2026 and H1 2026 filings. Non-GAAP definitions differ by company and are not comparable across rows.
Why Is Xiaomi’s EV Unit Still Losing Money at 100,000 Cars a Quarter?
Because the segment it sits in is paying for more than cars, and Xiaomi has not broken it out.
Xiaomi delivered 104,199 vehicles in the second quarter of 2026, up 28.2% year on year, with the segment labelled “smart EV, AI and other new initiatives” generating RMB 24.9 billion of revenue and an operating loss of RMB 2.6 billion. Segment gross margin fell to 19.2% from 26.4%. Revenue per vehicle slipped about 9.6% to roughly RMB 229,000, which the company put down to a lower mix of the high-end SU7 Ultra.
Divide the loss by the cars and you get RMB 24,952 per vehicle. That figure appears in our dataset, and it appears with a warning attached.
Axis Intelligence Research does not rank Xiaomi in the PPV table, and the reason matters. The RMB 2.6 billion is a segment operating loss covering artificial intelligence and other new initiatives alongside vehicle manufacturing — RMB 2.4 billion of the group’s RMB 3.6 billion quarterly capital expenditure went into that same segment. It is not net profit attributable, and the denominator does not cover everything the numerator paid for. Placing it in a column next to Li Auto’s RMB −20,642 would suggest a comparison that the disclosure does not support.
This is the kind of judgement that separates a dataset from a spreadsheet. The number is published, sourced and dated. It is simply not the same measurement, and we say so rather than letting the table imply otherwise.
Aidan Jad: Xiaomi’s EV arm is roughly two years into volume production and running a per-vehicle deficit larger than Li Auto’s after six years. That comparison is unfair in both directions and it is still the one everyone will draw. What the segment reporting genuinely shows is a company spending two-thirds of its group capital expenditure on a business unit that is one-quarter of its revenue. Whether that is a loss or an investment depends on 2028, not on this quarter’s income statement.
How Are Exports Changing Chinese EV Company Economics?
Overseas volume is now the variable that most separates these companies’ margins, and the disclosures have started to say so explicitly.
BYD reported overseas sales of about 792,000 passenger vehicles and pickups in the half, up roughly 68%, and — the disclosure that will define the reporting season — overseas revenue reached 52.57% of group revenue, the first period in which international sales exceeded domestic. Geely’s exports rose 158% to more than 474,000 units, already above its full-year 2025 total of 420,000, prompting the company to raise its 2026 export target from 640,000 to 920,000. Leapmotor’s exports rose 372.6% to 96,294 units, or 27.0% of everything it delivered.
| Company | H1 2026 export or overseas volume | Change YoY | Share of deliveries | Source |
|---|---|---|---|---|
| BYD | ~792,000 | +68% | 43.8% | BYD 2026 Interim Report |
| Geely Automobile | 474,000 | +158% | 33.3% | Geely Automobile Holdings |
| Leapmotor | 96,294 | +372.6% | 27.0% | Leapmotor interim results |
Scopes differ: BYD reports overseas sales of passenger vehicles and pickups, Geely and Leapmotor report export volume. Share of deliveries calculated by Axis Intelligence Research against each company’s own delivery base.
The pattern is not subtle. The three profitable companies in this dataset are the three with the largest export books, and the three deepest per-unit losses belong to companies whose volume is overwhelmingly domestic. That is correlation, and the causal mechanism is well documented in the pricing: vehicles sold abroad carry higher transaction prices than the same models sold into China’s discount cycle.
For the destination-level customs data behind those volumes, and the export dependence readings that go with them, see our China EV export statistics analysis. The national fleet, charging and battery context sits in China EV statistics, and company-level detail on the market leader is in BYD statistics.
A second calculation Axis Intelligence Research declines to publish: we do not construct a per-vehicle profit split between domestic and export sales for any company here. None of the seven discloses profit by geography at the vehicle level. Estimates of an overseas margin premium circulate widely; they are analyst inference, not company disclosure, and this dataset carries only what the filings say.
What the Second Quarter Changed
Reading the quarterly sequence rather than the half-year totals shows two companies moving in opposite directions inside the same market.
| Company | Q1 2026 net result | Q2 2026 net result | Direction |
|---|---|---|---|
| NIO | −RMB 496.0m | −RMB 721.6m | Widened |
| XPeng | −RMB 1,784.1m | −RMB 1,337.1m | Narrowed 25.1% |
| Li Auto | −RMB 2,289.5m | −RMB 1,704.2m | Narrowed 25.6% |
Net loss attributable to ordinary shareholders. Source: company Q2 2026 filings.
Both Li Auto and XPeng cut their quarterly losses by about a quarter. NIO’s widened in absolute terms even as its deliveries rose 29.0% quarter on quarter and its adjusted operating line stayed positive at RMB 206.9 million — the gap sits in interest expense, equity-investee losses and accretion on redeemable non-controlling interests, none of which appear in an operating margin.
Research spending tells its own story about how each company is playing the next two years. XPeng raised R&D 32.1% year on year to RMB 2.91 billion in the quarter. NIO cut it 28.7% to RMB 2.14 billion, on what it described as organizational optimization. Li Auto held roughly flat at RMB 2.78 billion. BYD spent RMB 28.9 billion in the half — more than its entire net profit attributable for the period.
Mia Scarlett: A company spending more on research than it earns is either building the next decade or losing control of its cost base, and the income statement alone will not tell you which. What it will tell you is that BYD chose to do this in the half its profit fell 20.5%, its volumes fell 15.7%, and it declined to declare an interim dividend. That is a set of decisions, not a set of results.
Methodology
Collection. Every figure on this page was retrieved from a filing or issuing-organisation release during production on September 1, 2026, and logged with its URL and retrieval date in the accompanying CSV. Li Auto, NIO and XPeng figures come from their second-quarter 2026 results filings, read directly — Li Auto and NIO through Form 6-K exhibits on SEC EDGAR, XPeng through its investor relations release. Geely figures come from the company’s own H1 2026 results release. BYD, Leapmotor and Xiaomi figures come from those companies’ interim and quarterly results announcements.
Formulas.
- PPV (Profit Per Vehicle) = net profit attributable to shareholders ÷ vehicles delivered, same disclosure, same period. Reported in RMB per vehicle. Where a company also publishes a core or adjusted profit line, PPV is shown on both bases with each basis named.
- UBG (Unit Breakeven Gap) = |net loss attributable to shareholders| ÷ gross profit per vehicle, same period. Gross profit per vehicle = group gross profit ÷ vehicles delivered. Holds operating expenses and unit gross profit constant.
- Segment operating loss per vehicle (Xiaomi only) = segment operating loss ÷ vehicles delivered, same quarter. Reported separately from PPV and not ranked alongside it.
Every calculation was run twice, in floating-point and exact rational arithmetic, and cross-checked before publication. All inputs are published in the CSV, so any reading here can be recomputed independently.
Definitions and basis. Six measurement conventions appear on this page and are never merged. BYD’s volume is new energy wholesale. Geely’s is all vehicles including combustion. Li Auto, NIO and XPeng report deliveries to customers. Xiaomi reports smart EV segment deliveries. Leapmotor reports global deliveries including exports. Profit figures are net profit attributable to shareholders of the parent unless a row states otherwise. Where a half-year figure is not filed as such, it is the sum of the two filed quarters, and the method note in the CSV says so.
Source-grade flagging. Rows carry is_primary = yes where the figure was read directly from the issuing organisation’s own filing or release during production, and is_primary = no where the figure originates in that organisation’s filing but was obtained from reporting of it rather than from the document itself. Of 172 rows, 131 are primary-read. Every metric in the two Axis rankings — PPV and UBG — rests on primary-read rows for Li Auto, NIO, XPeng and Geely.
Disclosed measurement differences. Documented rather than reconciled: Geely’s core profit and net profit attributable differ by RMB 590 million on foreign exchange and impairment, and both are shown. NIO’s statutory net loss and non-GAAP net profit differ by RMB 929.6 million of share-based compensation, and both are shown. Xiaomi’s per-vehicle figure is a segment operating loss and is excluded from the PPV ranking. Non-GAAP definitions are set by each company individually and are not comparable across companies.
About This Dataset
chinese-ev-companies.csv contains 172 observations covering deliveries, revenue, gross profit, net profit, margins, export volumes, cash positions, research spending and guidance for seven listed Chinese EV manufacturers, spanning the first half of 2025 through the third-quarter 2026 guidance issued in August and September 2026. Every row carries the issuing organisation, source document, source URL, retrieval date, whether the figure was read from the primary document, whether the value is calculated by Axis Intelligence Research, and — for calculated rows — the full formula.
License: CC BY 4.0. Free to use, including commercially, with attribution.
Citation: Axis Intelligence Research, Chinese EV Companies 2026, 2026.
Frequently Asked Questions
Which Chinese EV company makes the most profit per vehicle?
BYD, at RMB 6,815 per new energy vehicle in the first half of 2026 on Axis Intelligence Research’s calculation, narrowly ahead of Geely at RMB 6,803 on a core profit basis. The two are close enough that the ranking between them depends on which profit line you use: on net profit attributable, Geely falls to RMB 6,388 because of foreign exchange movements, while BYD’s figure is unchanged. Both are an order of magnitude above the only profitable startup in the group.
Is Leapmotor really profitable, or is it an accounting result?
It is a statutory result. Leapmotor reported net profit attributable to equity holders of RMB 210 million for the first half of 2026, its third consecutive profitable half-year, with positive operating cash flow and RMB 38.59 billion of cash on hand. The qualification is scale rather than accounting: RMB 210 million across 356,487 vehicles is RMB 589 per car and a 0.55% net margin, which leaves almost no absorption capacity if raw material costs move again. Gross margin already fell from 14.1% to 11.7% over the year.
Why did BYD’s profit fall when it is still the market leader?
Three things happened at once. Volume fell 15.72% as China’s domestic NEV market contracted, revenue fell 7.13% to RMB 344.8 billion, and research spending of RMB 28.9 billion exceeded net profit for the period. Per-unit profitability held up far better than the headline suggests: PPV fell only 5.71%, from RMB 7,228 to RMB 6,815. BYD lost volume, not pricing discipline — and its second-quarter net profit rose about 30% year on year even as the half-year total fell 20.5%.
How is Profit Per Vehicle different from gross margin or vehicle margin?
Gross margin tells you what a company keeps after building the car. PPV tells you what is left after building it, selling it, developing the next one, servicing debt and paying tax. XPeng illustrates the gap: its second-quarter gross margin of 20.7% was the highest of the three U.S.-listed startups, and its PPV was negative RMB 18,805. A healthy gross margin on a cost base that is too large for the volume still produces a loss on every car.
Can Li Auto, NIO or XPeng grow their way to profitability?
The Unit Breakeven Gap puts numbers on the question. NIO would need about 21,616 additional vehicles per half-year at its current gross profit of RMB 56,328 per car — roughly 11% more volume, and its third-quarter guidance of 108,000 to 111,000 vehicles is in that territory. XPeng would need 76,565, or 46% more. Li Auto would need 166,377, or 86% more, against third-quarter guidance that is roughly flat. Same industry, same price war, three completely different distances to the line.
Does Xiaomi lose more money per car than the established EV startups?
The comparison does not hold as usually stated. Xiaomi’s RMB 2.6 billion second-quarter loss belongs to a segment covering smart EV, artificial intelligence and other new initiatives, and RMB 2.4 billion of group capital expenditure went into that same segment in the quarter. Dividing it by 104,199 vehicles gives RMB 24,952, but the numerator is paying for work that has nothing to do with the cars. Xiaomi has not published a net result for the vehicle business alone, so no like-for-like figure exists.
Why do Chinese EV companies report so many different profit figures?
Because each is measuring a different thing, and none of them is wrong. Geely publishes core profit to strip out currency swings that have nothing to do with selling cars. NIO publishes adjusted net profit to exclude share-based compensation, which is non-cash. XPeng adjusts for both share-based compensation and a derivative fair-value gain tied to an acquisition. The figures diverge sharply — NIO’s statutory loss of RMB 1.22 billion becomes a RMB 69.6 million profit on the adjusted line. Quoting one without naming its basis is what creates the confusion.
What would change these rankings in the second half of 2026?
Four things are visible in the disclosures. Export mix, which is already the clearest dividing line between the profitable and the unprofitable. Product cycle timing — Li Auto refreshed its entire L series in the first half, which compressed margin, and management expects mix to improve. Raw material costs, which both Leapmotor and Xiaomi named as the reason their gross margins fell. And the domestic price war, which set the ceiling on every transaction price in this dataset and has not shown a sign of easing.
Are these figures audited?
No, and that is normal for interim reporting. Quarterly results from Li Auto, NIO and XPeng are unaudited, though reviewed by the companies’ audit committees and, in some cases, their independent auditors. Hong Kong interim results are unaudited but subject to auditor review. Full-year figures for 2026 will be audited and may be restated. Every row in the accompanying dataset carries its as-of date so that a later revision can be identified rather than silently absorbed.
Cite This Research
APA — Axis Intelligence Research. (2026). Chinese EV companies 2026: Sales, profit per vehicle and who actually makes money. https://axis-intelligence.com/chinese-ev-companies/
MLA — Axis Intelligence Research. “Chinese EV Companies 2026: Sales, Profit Per Vehicle and Who Actually Makes Money.” Axis Intelligence, 1 Sept. 2026, axis-intelligence.com/chinese-ev-companies/.
Chicago — Axis Intelligence Research. “Chinese EV Companies 2026: Sales, Profit Per Vehicle and Who Actually Makes Money.” Axis Intelligence, September 1, 2026. https://axis-intelligence.com/chinese-ev-companies/.
The China EV Research Cluster
- China EV Statistics 2026 — fleet, charging, batteries and the 2030 target
- China EV Export Statistics 2026 — destinations, export dependence and the substitution shift
- BYD Statistics 2026 — revenue per vehicle, energy storage and the gap with Tesla
- Tesla Statistics 2026 — the benchmark these companies are measured against
- Electric Vehicle Statistics 2026 — the global picture
- Electric Vehicles Index — real-world range and charging-curve data