Xiaomi EV Sales Statistics 2026
By Axis Intelligence Research
Co-author: Aidan Jad, EV & Clean Energy | Last updated: September 20, 2026 | License: CC BY 4.0
Xiaomi delivered 246,475 vehicles between January and August 2026 — 44.8% of its 550,000-unit full-year target with four months left on the clock. To reach it, the company must average 75,881 deliveries a month from September through December, 2.37 times its June–August run-rate. That multiple is the single number that governs Xiaomi’s automotive year.
Quick Answer
Xiaomi Auto delivered 246,475 vehicles in the first eight months of 2026, against a stated full-year target of 550,000. Average selling price fell to RMB229,312 in Q2 2026, below Xiaomi’s own 2024 full-year average. The Axis Delivery Gap Index (DGI) reads 2.37 as of August 31, 2026 — the required run-rate is 2.37× the current one.
Key Findings
- Axis Intelligence Research finds Xiaomi delivered 246,475 vehicles in January–August 2026, equal to 44.8% of its 550,000-unit target, based on CPCA monthly data and company filings.
- The Axis Delivery Gap Index for Xiaomi reads 2.37 as of August 31, 2026, the widest reading since the company began publishing annual delivery targets.
- Xiaomi’s smart EV average selling price fell to RMB229,312 in Q2 2026, down 9.6% year-over-year and 11.8% below its Q3 2025 peak of RMB260,053.
- Axis Intelligence Research estimates Xiaomi’s smart EV, AI and new-initiatives segment carried an operating loss of roughly RMB24,950 per vehicle delivered in Q2 2026, down 35% from Q1.
- Xiaomi’s Q2 2026 deliveries grew 28.2% while China’s passenger vehicle retail market contracted 22%, a 50-point relative gap, per Xiaomi’s Q2 2026 results announcement citing CPCA.
How Many Cars Has Xiaomi Sold in 2026?
Through August 31, 2026: 246,475 units. That figure combines the 216,322 vehicles Xiaomi delivered in January–July with 30,153 in August, both on China Passenger Car Association retail basis.
The shape of the year matters more than the total. January was the strongest month, above 39,000 units on YU7 momentum. Spring Festival pulled February and March back into the low twenty-thousands. From April onward Xiaomi cleared 30,000 units for five consecutive months — and then stopped growing.
Xiaomi EV deliveries by quarter, 2024–2026
| Period | Deliveries (units) | Source |
|---|---|---|
| FY2024 | 139,487 | Xiaomi FY2025 annual results announcement |
| Q1 2025 | 75,869 | Xiaomi Q1 2026 results announcement |
| Q2 2025 | 81,302 | Xiaomi Q2 2026 results announcement |
| Q3 2025 | 108,796 | Xiaomi Q3 2025 results announcement |
| Q4 2025 | 145,115 | Xiaomi FY2025 annual results announcement |
| FY2025 | 411,082 | Xiaomi FY2025 annual results announcement |
| Q1 2026 | 80,856 | Xiaomi Q1 2026 results announcement |
| Q2 2026 | 104,199 | Xiaomi Q2 2026 results announcement |
| H1 2026 | 185,055 | Axis Intelligence Research (Q1 + Q2) |
Xiaomi EV monthly deliveries, mid-2026
| Month | Deliveries (units) | YoY change | Source |
|---|---|---|---|
| Jan–May 2026 (cumulative) | 150,317 | — | CPCA |
| June 2026 | 34,738 | — | Axis Intelligence Research (H1 less Jan–May) |
| July 2026 | 31,267 | +2.7% | CPCA |
| August 2026 | 30,153 | −17.2% | CPCA |
| July–August 2026 | 61,420 | −8.1% | Axis Intelligence Research |
That last row is the one nobody is publishing. Taken as a pair, Xiaomi’s July and August volume came in 8.1% below the same two months of 2025 — the summer window when a plant running double shifts should be at its most productive. Q4 2025 alone produced 145,115 units. The whole of Q3 2026 so far has produced 61,420.
Aidan Jad: The instinct is to read a soft August as demand softening, and the order bank says otherwise — the YU7 was still quoting multi-month waits into the summer. What actually happened is a changeover. The new-generation SU7 rolled out in March, the first-generation line was discontinued, and Phase 2 spent the period being validated rather than pushed. A plant in the middle of a model changeover does not produce its own peak. The question is whether the line comes back at the higher takt rate the target assumes.
Will Xiaomi Hit Its 550,000 Delivery Target in 2026?
On the current trajectory, no — and the arithmetic is not close.
Xiaomi confirmed the 550,000 figure in its FY2025 annual results announcement, stating it would “strive to achieve the target of delivering 550,000 vehicles for the entire year.” After eight months, 303,525 units remain. Spread over September–December, that is 75,881 vehicles a month — against an August print of 30,153.
This is the gap the Axis Delivery Gap Index measures.
The Axis Delivery Gap Index (DGI)
DGI = required monthly run-rate for remaining months ÷ trailing three-month average monthly deliveries
Both inputs are public, both are dated, and any analyst can recompute the reading from the two numbers we publish.
| Input | Value | Basis |
|---|---|---|
| Full-year target | 550,000 units | Xiaomi FY2025 annual results announcement |
| Delivered Jan–Aug 2026 | 246,475 units | CPCA monthly data, compiled |
| Remaining requirement | 303,525 units | Target less delivered |
| Months remaining | 4 | September–December |
| Required run-rate | 75,881 units/month | 303,525 ÷ 4 |
| Trailing 3-month average | 32,053 units/month | Mean of June, July, August |
| DGI reading | 2.37 | 75,881 ÷ 32,053 |
How to read a DGI reading. At or below 1.00, the target is already covered by the current pace. 1.00–1.25 is reachable on normal seasonal lift and a strong December push. 1.25–1.75 requires a genuine step change — a new nameplate ramping cleanly, or a shift added. Above 1.75, the target needs a discontinuity in output, not an acceleration of it.
Xiaomi’s 2.37 sits well outside that last band. For context on what 2.37 asks for: the company’s single best month on record came inside a Q4 that averaged roughly 48,000 units. The target now asks for 75,881 a month, four times running, on a line that has just added a third product series.
Full component definitions, the reading scale and the running history live on the Axis Delivery Gap Index page, which is updated as each month’s CPCA data lands.
Can the Beijing Plant Physically Build 550,000 Cars?
This is where the target stops being a demand question and becomes a manufacturing one.
Xiaomi builds every car it sells at a single site in Beijing’s Yizhuang district. Under the November 2021 agreement with the Beijing Economic-Technological Development Area, the plant was specified as two phases of 150,000 units a year each — 300,000 units of nameplate capacity. Phase 2 entered mass production during 2025. A Xiaomi subsidiary secured a 485,133.91 m² adjacent plot in June 2025 for RMB635 million, the site widely understood to be earmarked for Phase 3; it is not yet producing vehicles.
Set output against that nameplate and the picture sharpens. Axis Intelligence Research calls this ratio the Nameplate Output Ratio (NOR) — deliveries divided by pro-rated nameplate capacity.
| Period | Deliveries | Pro-rated nameplate | NOR | Basis |
|---|---|---|---|---|
| FY2025 | 411,082 | 300,000 | 1.37 | Axis Intelligence Research |
| Jan–Aug 2026 | 246,475 | 200,000 | 1.23 | Axis Intelligence Research |
| FY2026 at target | 550,000 | 300,000 | 1.83 | Axis Intelligence Research |
Xiaomi is already building 23% more cars than its nameplate says it can, and built 37% more than nameplate across 2025. That is double-shift territory and it is genuinely impressive throughput. It also means the 550,000 target does not ask Xiaomi to fill spare capacity — there is none. It asks the same two phases to run at 1.83× nameplate for a full year, a rate no phase of the plant has yet sustained over twelve months.
Aidan Jad: Every ramp story eventually runs into the same wall, and it is never the order book. It is fixturing, line balance and supplier takt. Xiaomi’s numbers say the Beijing plant is one of the hardest-worked assembly lines in China right now. You do not get another 60% out of a line running at 123% of nameplate by asking the team to try harder — you get it from Phase 3 pouring concrete, or you do not get it.
Xiaomi EV Revenue and ASP: Why Revenue Is Growing Slower Than Volume
Volume and revenue have decoupled. Q2 2026 deliveries rose 28.2% year-over-year; smart EV revenue rose 15.9%. The 12-point wedge between them is price.
Xiaomi smart EV average selling price by quarter
| Period | ASP (RMB/unit) | Source |
|---|---|---|
| FY2024 | 234,479 | Xiaomi FY2025 annual results announcement |
| Q1 2025 | 238,301 | Xiaomi Q2 2025 results announcement |
| Q2 2025 | 253,662 | Xiaomi Q2 2026 results announcement |
| Q3 2025 | 260,053 | Xiaomi Q3 2025 results announcement |
| Q4 2025 | 249,846 | Xiaomi Q1 2026 results announcement |
| FY2025 | 251,171 | Xiaomi FY2025 annual results announcement |
| Q1 2026 | 235,116 | Xiaomi Q1 2026 results announcement |
| Q2 2026 | 229,312 | Xiaomi Q2 2026 results announcement |
Q2 2026’s RMB229,312 is not just a decline. It is below Xiaomi’s full-year 2024 average of RMB234,479 — the company’s very first year of deliveries, before the SU7 Ultra or the YU7 existed. Eighteen months of premiumisation has round-tripped. Xiaomi is explicit about the cause in its Q2 2026 results announcement: a lower contribution from the RMB500,000-plus SU7 Ultra, and vehicle purchase-tax subsidies absorbed by the company.
Smart EV revenue
| Period | Smart EV revenue (RMB bn) | Source |
|---|---|---|
| FY2024 | 32.1 | Xiaomi FY2025 annual results announcement |
| FY2025 | ~103.3 | Axis Intelligence Research (411,082 × RMB251,171) |
| Q2 2025 | 20.6 | Xiaomi Q2 2026 results announcement |
| Q1 2026 | 19.0 | Xiaomi Q2 2026 results announcement |
| Q2 2026 | 23.9 | Xiaomi Q2 2026 results announcement |
The wider smart EV, AI and other new initiatives segment reached RMB24.90 billion in Q2 2026, 22.9% of group revenue and up 17.1% year-over-year. That segment crossing 20% of a company that still sold 31.2 million smartphones in the quarter is the structural story investors keep coming back to.
Is Xiaomi Making Money on Its Cars Yet?
Not at the segment line. Xiaomi’s smart EV, AI and other new initiatives segment posted an operating loss of RMB2.6 billion in Q2 2026 on a 19.2% gross margin — down from 26.4% a year earlier.
Axis Intelligence Research converts that into a per-vehicle scale indicator, dividing the segment’s operating result by deliveries in the same period:
| Period | Segment operating loss (RMB m) | Deliveries | Loss per vehicle (RMB) |
|---|---|---|---|
| Q1 2026 | −3,106 | 80,856 | −38,415 |
| Q2 2026 | −2,600 | 104,199 | −24,952 |
| H1 2026 | −5,714 | 185,055 | −30,878 |
Q1 and H1 operating losses are derived from the filed segment gross profit less the filed segment operating expenses: 3,993.9 − 7,100 = −3,106.1 for Q1; 8,785.9 − 14,500 = −5,714.1 for H1. Q2’s −2,600 is stated directly in the filing.
Read this as a scale ratio, not a bill of materials. The segment carries Xiaomi’s foundation-model work and other new initiatives alongside the cars, so the true per-car number is better than these figures. What the series shows cleanly is direction: the loss per delivered vehicle narrowed 35% between Q1 and Q2 2026, because 23,343 extra units absorbed a fixed-cost base that grew only 4.5%. Operating leverage is arriving on schedule. Price is taking it away again.
Capital intensity tells the same story. Segment capex ran RMB3.88 billion in H1 2026 — RMB20,949 per vehicle delivered, by Axis Intelligence Research’s calculation, on a business still building its second and third production phases.
Aidan Jad: Margin compression from 26.4% to 19.2% in four quarters looks alarming until you separate what caused it. Roughly speaking: mix, components, and AI cost allocation. Only one of those is about the cars. Pack costs and the memory squeeze are industry-wide and cyclical; the Ultra mix reverting toward the mean was always going to happen once the halo model’s launch cohort cleared. The number worth tracking is not the margin, it is whether ASP stops falling now that SkyNomad has arrived at a lower price point than the SU7.
Which Xiaomi Model Sells Best: SU7 or YU7?
The SUV has taken the lead, and it happened quietly.
Across January–July 2026, the SU7 sedan series delivered 101,540 units on CPCA data. Total deliveries for the same period were 216,322. That puts the YU7 series at 114,782 units — 53.1% of Xiaomi’s 2026 volume, by Axis Intelligence Research’s calculation.
| Model line | Jan–Jul 2026 (units) | Share | July 2026 (units) |
|---|---|---|---|
| SU7 series (incl. Ultra) | 101,540 | 46.9% | 21,044 |
| YU7 series | 114,782 | 53.1% | 10,223 |
| Total | 216,322 | 100% | 31,267 |
YU7 figures calculated by Axis Intelligence Research as total less SU7 series. Source: CPCA monthly data.
The monthly picture inverts that, and it is the more useful signal: July saw the SU7 deliver 21,044 against the YU7’s 10,223, with the SUV down 28.6% sequentially. A model that carried more than half the year’s volume and then halved its monthly rate is not a demand story — it is a line-allocation story, and it points at the same Phase 2 changeover the quarterly data implies.
Cumulative SU7 series deliveries passed 500,000 units on August 17, 2026, per Xiaomi’s Q2 filing. Xiaomi Auto’s all-model cumulative total passed 800,000 on September 7, 2026.
On residuals — the number that actually predicts repeat purchase in China — Xiaomi is leading. The July 2026 China Vehicle Value Retention Ratio Research Report from the China Automobile Dealers Association and Jingzhengu placed the YU7 first among one-year major battery-electric vehicles at 82.8%, with the SU7 series fourth at 75.3% and first among one-year BEV sedans.
What Does the SkyNomad Launch Change?
Everything, or nothing, depending on the ramp.
Xiaomi launched the SkyNomad series on September 7, 2026 — its third product line and its first move out of pure battery-electric into extended-range. Deliveries started September 12–13 across 75 Chinese cities.
| Variant | Launch price (RMB) | Pre-sale price (RMB) | Delta | Layout |
|---|---|---|---|---|
| SkyNomad N70 Pro | 209,900 | not pre-sold | — | 5-seat, RWD |
| SkyNomad N70 Max | 239,900 | 259,900 | −20,000 | 5-seat, AWD |
| SkyNomad N90 Max | 269,900 | 299,900 | −30,000 | 7-seat, AWD |
| SkyNomad N90 Max Explorer | 299,900 | not pre-sold | — | pop-up roof |
Pre-sale prices per Xiaomi’s Q2 2026 results announcement; launch prices per the September 7, 2026 launch event. Deltas calculated by Axis Intelligence Research.
Two of the four launched below their own pre-sale prices — a 7.7% cut on the N70 Max and a 10% cut on the N90 Max. Pre-sale-to-launch price cuts of that size are a competitive read, not a generosity: they say Xiaomi looked at the EREV segment between RMB200,000 and RMB300,000 and decided its July pricing would not clear volume. Xiaomi reported more than 10,000 locked orders in the first four minutes.
The segment it entered is shrinking. CPCA retail data puts Chinese EREV sales at roughly 535,000 units in January–July 2026, down 15.35% year-over-year. Xiaomi has launched its highest-volume-intent product into the one powertrain category in China that is contracting.
The N90 Max carries a 1,705 km CLTC combined range on a 76 kWh pack with a 1.5-litre range extender, per Xiaomi’s filing.
Aidan Jad: Price the entry car at 209,900 and you have changed your own ASP math permanently. Every SkyNomad Pro that ships instead of a YU7 pulls the blended ASP down by roughly a third of the gap between them. That is fine — it is what a volume play looks like — but it means the RMB229,312 ASP is not the floor. Watch the Q3 print for the first clean reading, and watch whether locked orders convert inside the one-to-five-week window Xiaomi quoted, because an EREV ramp with a new engine supplier is exactly where launch cadence slips.
How Does Xiaomi Compare With the Chinese Market Overall?
Xiaomi is growing while the market it sells into is not.
In Q2 2026, China’s overall passenger vehicle retail volume fell 22% year-over-year on CPCA data, as cited in Xiaomi’s own results announcement. Xiaomi’s deliveries rose 28.2% in the same quarter. Axis Intelligence Research puts the relative gap at 50.2 percentage points — the widest outperformance of any quarter Xiaomi has reported.
That context reframes the target discussion. The 550,000 figure was set in early January 2026, before the market contracted this hard. Judged against the plan, Xiaomi is behind. Judged against the market, it is one of the few nameplates still adding volume.
Retail network: 477 Xiaomi EV sales centres across 138 cities in the Chinese Mainland as of December 31, 2025.
When Does Xiaomi Start Selling Cars in Europe?
- Xiaomi Auto confirmed European entry on September 3, 2026, signing memoranda of understanding with the first eight leading German dealer groups at IFA Berlin 2026.
The channel choice is the interesting part. Xiaomi is not replicating the direct-sales model most Chinese entrants and Tesla have used in Europe — it is buying into Germany’s incumbent franchised retail infrastructure from day one. That is a more expensive route to a service network and a materially faster one. None of that volume lands in 2026, which is why it does not move the DGI reading, but it does define what the Beijing capacity question looks like in 2027 and beyond.
Methodology
Collection. Quarterly delivery, revenue, ASP, margin, capex and segment figures were taken directly from Xiaomi Corporation’s filed results announcements (HKEX: 1810) for Q1 2026, Q2/H1 2026, Q3 2025 and FY2025, each fetched and read on September 20, 2026. Monthly delivery figures are China Passenger Car Association retail data as published in its monthly releases. Product pricing and launch dates are company-announced. Every figure in this article appears as a row in the downloadable dataset with its source organisation, document, URL where one exists, and retrieval date.
Two reporting bases. Xiaomi reports quarterly deliveries in its filings; CPCA reports monthly retail. The two series agree closely but are not identical in definition. Quarterly analysis in this article uses filed figures; monthly analysis uses CPCA. Where the two are combined — as in the January–August total — the basis is stated in the dataset’s method_note column.
Axis Delivery Gap Index (DGI). DGI = (full-year target − year-to-date deliveries) ÷ months remaining, divided by the trailing three-month average monthly delivery rate. Inputs for the September 20, 2026 reading: target 550,000 (company-stated); year-to-date 246,475 through August 31; four months remaining; trailing three-month average 32,053 (June 34,738, July 31,267, August 30,153). Reading: 75,881 ÷ 32,053 = 2.37. The index is unweighted by design — a weighted version would embed a seasonality assumption and make the reading unreproducible from public data. Readings are snapshot-dated and revised only when an input is restated at source; prior readings are never rewritten.
Nameplate Output Ratio (NOR). NOR = deliveries ÷ pro-rated nameplate annual capacity. Nameplate is taken as 300,000 units (Beijing Yizhuang Phases 1 and 2 at 150,000 each, per the November 2021 development-area agreement). NOR measures output against the officially specified figure, not against instantaneous line capability, which Xiaomi does not disclose. A reading above 1.00 indicates output above nameplate.
Segment loss per vehicle. Operating loss of the smart EV, AI and other new initiatives segment ÷ deliveries in the same period. The segment includes non-automotive activity, so the ratio is a scale indicator of segment economics rather than a per-unit vehicle margin, and it overstates the loss attributable to vehicles.
What this dataset does not cover. Xiaomi does not disclose vehicle-only operating profit, per-model ASP, production volume as distinct from deliveries, or order-bank size. Those figures are absent here rather than estimated.
About This Dataset
Contents. 80+ observations covering Xiaomi Auto deliveries (annual, quarterly and monthly, 2024–2026), smart EV ASP and revenue by quarter, segment gross margin, operating result, capex, model-line splits, SkyNomad pricing, nameplate capacity, residual values and Chinese market context — each row carrying its source organisation, document, URL, retrieval date and a primary/secondary flag.
License. CC BY 4.0. Free to reuse, republish and build on with attribution.
Cite as: Axis Intelligence Research, Xiaomi EV Sales Statistics 2026: Deliveries, ASP, Margins and the 550,000 Target Gap, 2026.
Citation Block
APA — Axis Intelligence Research. (2026). Xiaomi EV sales statistics 2026: Deliveries, ASP, margins and the 550,000 target gap. https://axis-intelligence.com/xiaomi-ev-sales-statistics/
MLA — Axis Intelligence Research. “Xiaomi EV Sales Statistics 2026: Deliveries, ASP, Margins and the 550,000 Target Gap.” Axis Intelligence, 2026, axis-intelligence.com/xiaomi-ev-sales-statistics/.
Chicago — Axis Intelligence Research. “Xiaomi EV Sales Statistics 2026: Deliveries, ASP, Margins and the 550,000 Target Gap.” Axis Intelligence, 2026. https://axis-intelligence.com/xiaomi-ev-sales-statistics/.
Frequently Asked Questions
What is the Axis Delivery Gap Index and how do I recompute it?
DGI divides the monthly delivery rate an automaker needs for the rest of its target year by its trailing three-month average delivery rate. For Xiaomi on August 31, 2026: (550,000 − 246,475) ÷ 4 = 75,881 required, against a June–August average of 32,053, giving 2.37. Both inputs are public, so any reading we publish can be checked against CPCA data in under a minute.
Does Xiaomi’s monthly “over 30,000 deliveries” announcement mean the same thing as the CPCA figure?
No, and the difference matters when you are modelling. Xiaomi announces a floor — it has used the phrase “continued to exceed 30,000 units” for five consecutive months without disclosing a figure or a model split. The exact number and the SU7/YU7 breakdown arrive later from CPCA retail data. August’s actual print was 30,153, a 17.2% year-over-year decline that the company announcement did not convey.
Why did Xiaomi’s smart EV gross margin fall from 26.4% to 19.2% in a year?
Three causes, per Xiaomi’s Q2 2026 filing: a lower share of the high-ASP SU7 Ultra in the delivery mix, higher prices for key components, and AI-business costs booked to the same segment. Only the first is about vehicle economics. The segment reports EVs, foundation-model work and other new initiatives together, so segment margin is not vehicle margin.
Is the 550,000 target constrained by demand or by factory capacity?
By capacity, on the evidence. Xiaomi delivered 246,475 units in eight months against a pro-rated nameplate of 200,000 — a Nameplate Output Ratio of 1.23, meaning the Beijing plant is already running well above its specified rate. Hitting 550,000 would require an NOR of 1.83 across the full year. The YU7’s quoted wait times through the summer point the same way: order intake is not the binding constraint.
What does SkyNomad have to sell for Xiaomi to reach the target?
Roughly 45,700 units a month on top of a flat SU7/YU7 base of about 30,000 — for four consecutive months, from a standing start on September 12. No Chinese nameplate has ramped a new series to that rate inside a quarter. A more useful question for Q4 modelling is whether SkyNomad lifts the combined run-rate toward 40,000, which the first-week order pace makes plausible.
Is the SkyNomad an EREV or a plug-in hybrid, and why does it matter for these numbers?
It is an extended-range electric vehicle: the 1.5-litre engine generates electricity and never drives the wheels. That places it in a category whose Chinese retail volume fell 15.35% year-over-year to roughly 535,000 units in January–July 2026, per CPCA — so Xiaomi is adding capacity into a contracting powertrain segment, which is the main risk to the Q4 ramp thesis.
How do Xiaomi’s residual values compare with other Chinese EVs?
The YU7 series ranked first among one-year major battery-electric vehicles at 82.8% value retention in the July 2026 China Vehicle Value Retention Ratio Research Report from the China Automobile Dealers Association and Jingzhengu. The SU7 series ranked fourth overall at 75.3% and first among one-year BEV sedans. Strong residuals lower the effective cost of ownership and support lease and financing volume — a quiet advantage over newer nameplates.
When will Xiaomi EVs be available outside China?
2027, starting with Germany. Xiaomi Auto signed MoUs with eight German dealer groups at IFA Berlin on September 3, 2026. No export volume is expected in 2026, so European plans have no bearing on the current-year delivery math — they bear on whether Phase 3 of the Beijing plant is a 2027 necessity rather than an option.
Where does Xiaomi’s EV business sit inside the wider group?
The smart EV, AI and other new initiatives segment generated RMB24.90 billion in Q2 2026, 22.9% of Xiaomi’s RMB108.9 billion group revenue, up from 18.3% a year earlier. It is the only segment growing: smartphone × AIoT revenue fell 11.3% year-over-year in the same quarter.
Corrections. Spotted an error? Write to [email protected]. Our corrections policy is public and every correction is dated on the page it affects.
Related research: Axis Delivery Gap Index · China EV Market Statistics · BYD Sales Statistics · Tesla China Sales Statistics · Li Auto Sales Statistics · EV Battery Statistics · XPeng Deliveries · Geely Sales Statistics