Geely's 500 Wh/kg battery, Chery's £69bn overseas push, and the week's most important numbers — 20 August 2026

On 20 August 2026, Electrek reported that Geely Holding — the parent of Geely, Zeekr, Lynk & Co, Volvo, Polestar, Lotus and Smart — has validated solid-state battery cells targeting 500 Wh/kg energy density, with a pilot deployment across those brands pencilled in for 2027.

That number deserves context before the headlines get ahead of it. The best high-nickel NMC cells in today's production EVs land around 250–300 Wh/kg at the cell level; Tesla's 4680 cell sits near 245 Wh/kg; and LFP, the chemistry powering most of the affordable Chinese EVs on UK roads, comes in lower still. At the pack level — once cells are put into modules and the full pack — energy density drops further. Zeekr's current Golden Brick LFP pack is rated around 128 Wh/kg at the pack level.

Doubling cell-level density to 500 Wh/kg is not an incremental gain. It means you can either keep today's range and cut battery mass roughly in half, or keep today's weight and roughly double the range. Geely is selling the range version of the trade: the company claims vehicles fitted with these cells could exceed 1,000 km per charge — comparable, it says, to a tank of diesel — and last up to one million kilometres.

Two specifics make this more credible than a typical OEM press release. First, Geely has named a materials partner: Dow Chemical, which developed an adhesive for the solid-state stack stable from −40°C to 120°C. Temperature stability is one of the hardest engineering problems in solid-state chemistry; a named supplier on that component is a real detail. Second, real-world validation runs started in 2026 — not in a laboratory, but on actual vehicles.

The honest caveat is equally important. A 2027 pilot is not mass production. A pilot line proves the cell can be built repeatably at low volume; scaling to the millions of units required for vehicle production, at a cost that works in a competitive market, typically takes several more years. For UK buyers, Geely's solid-state cell is a credible signal about the brand's long-term capability, not a reason to wait before buying a Zeekr or Polestar. It is, however, a data point that distinguishes Geely Holding from the field: the group has named a partner, started validation, and set a public timeline.

On 18 August 2026, BYD confirmed the specifications of the all-electric Denza N8 via China's Ministry of Industry and Information Technology (MIIT) new-vehicle approval filing — the most reliable source for pre-launch specs in the Chinese market. The numbers are striking: a 130.15 kWh battery providing 1,003 km of CLTC range, paired with BYD's Flash Charging technology capable of topping the car from 10% to 70% in five minutes and to 97% in nine.

For context, the UK's longest-range currently available Chinese EV — the BYD Seal U DM-i — is rated at approximately 420 miles WLTP on its hybrid system. The Denza N8 EV's 1,003 km is a CLTC figure, which is measured under Chinese testing conditions that tend to exceed real-world figures by 20–30%. Its probable real-world range would likely fall in the 700–800 km band — still roughly double what any Chinese EV sold in the UK today offers.

The Denza N8 is built on BYD's e3 platform and measures 5,150 mm long — a size class between a BMW 5-series and 7-series. It is designed for the Chinese luxury market; no UK launch has been announced and no RHD production is confirmed. The Denza sub-brand targets roughly the market segment that occupied the discontinued Denza D9 MPV, but the N8 is a five-seat large SUV. BYD appears to be broadening the Denza line to cover both formats.

Why does this matter to UK buyers? Two reasons. First, the Flash Charging claim — 10–70% in five minutes on a 130 kWh pack — implies a peak charge rate well above 1,000 kW. That is well beyond anything currently deployed in a UK public charging network; the UK's fastest public chargers currently peak at 350 kW. The implication is that, when BYD eventually brings Flash Charging-compatible vehicles to the UK, the infrastructure constraint will initially be on the network side, not the car. Second, the continuous movement of BYD's technical baseline upward signals that even the brand's UK-sold cars — which today cap at 150–230 kW — will eventually inherit these improvements.

Chery Auto (HKEX: 9973) reported its first-half 2026 results on 20 August, and the headline that UK buyers of Omoda and Jaecoo cars should understand is this: the Chinese car-maker behind both brands now derives 69% of its revenue from overseas markets — up from a lower base a year ago — and that share is growing rapidly.

The numbers break down as follows. Total revenue for H1 2026 was 143.28 billion yuan (approximately £15.7 billion), up 1.2% year-on-year — essentially flat in aggregate. But within that, overseas revenue rose 51% to 98.97 billion yuan, while internal-combustion vehicle revenue in the home market fell 24.8%, from 92.45 billion yuan to 69.51 billion yuan. The strategic direction is unambiguous: Chery is de-emphasising ICE in China and redirecting the business toward overseas sales and electrification.

The NEV share of Chery's revenue has more than doubled in twelve months, from 25.6% to 41.4%, on revenue of 59.28 billion yuan (up 63.8%). Gross margin improved meaningfully — from 13.0% to 16.1% — driven partly by the higher margins that overseas sales carry versus the brutally competitive Chinese domestic market.

The reason profit fell despite stronger gross margin is less about the core business and more about currency. Other income fell 27.1%, primarily due to lower foreign-exchange gains — a technical accounting effect rather than an operational deterioration. R&D spending rose 28.3% to 6.67 billion yuan, directed at electrification, new platforms, assisted driving, and smart cockpits. That is deliberate investment in future product, not a warning signal.

For UK buyers, the practical significance is about survival confidence. Chery is not a brand in distress; it is the largest Chinese car exporter, with 12 production bases globally including three overseas, and a UK R&D centre scheduled to open in Bedfordshire in late 2026. The Nissan Sunderland production announcement — Chery models assembled in the UK from 2027 — adds further local commitment. The profit dip is a currency effect and an R&D investment cost, not a sign of a brand struggling to make ends meet.

Two further stories from 20 August 2026 complete this week's picture of where the Chinese EV industry stands.

NIO (NYSE: NIO) will report its second-quarter 2026 results on 1 September 2026. The headline delivery figure — 107,658 vehicles in Q2, up 49.4% year-on-year — looks strong, but fell short of the 110,000–115,000 unit guidance range, with the shortfall concentrated in a weak April (29,356 deliveries). Q1 2026 delivered adjusted operating profit of 66.8 million yuan on revenue up 112.2% year-on-year; gross margin hit 19.0%, the highest in four years. Deutsche Bank expects Q2 profitability to hold, driven by higher-margin SUV models. NIO has no UK launch date, no RHD production, and no confirmed European commercial entity; it remains a brand to watch rather than to buy.

Meanwhile, Luxeed — the EV brand developed jointly by Huawei and Chery — opened pre-orders in China for its RX coupe SUV on 20 August, at 299,800 yuan (approximately £33,000 at current rates) for the standard variant with cabin LiDAR, and 359,800 yuan for an L3-ready variant. The designer is Werner Gruber, formerly of Ferrari (LaFerrari, FXX-K), which explains the low-slung proportions. The ADAS system is Huawei Qiankun, built on the Tuling platform. At these prices and with this hardware, Luxeed RX is a direct competitor to the Xiaomi YU7 in China's most crowded segment — premium sporty electric SUVs. For UK buyers: no RHD, no UK launch, no confirmed European rollout. It is an upstream intelligence signal about the Chery-Huawei partnership, which does supply Omoda and Jaecoo with Huawei connectivity technology.

On 18 August 2026, a Tesla Robotaxi operating without a safety monitor in Austin drove through a row of plastic bollards marking a closed lane. The video — posted to Reddit — showed the car stopping, creeping forward, reversing, and then simply pushing through the obstacle. No one was hurt. But the incident occurred less than a month after Tesla's head of AI told investors the Robotaxi programme had an 'impeccable safety record' with 'zero notable incidents.'

What is instructive here is not that Tesla's Robotaxi made a mistake — all ADAS systems have edge cases — but the nature of the failure. The car had clearly detected the bollards (it stopped for them). The failure was in the decision logic: committed to a right-turn lane that was closed, it could not work out how to exit the situation gracefully, so it proceeded through the obstacle. That is a planning failure, not a perception failure.

This matters for context when evaluating Chinese EV ADAS claims. The same week that Huawei was publicising its Guiyang ADAS competition results and BYD was promoting God's Eye across its range, Tesla's own driverless system — the only one with a significant operational commercial fleet outside China — hit a bollard and kept going. The lesson is consistent with what we report for UK buyers: no current mass-market ADAS system is reliably safe without human supervision, regardless of brand. The legal standard for road use in the UK requires a human driver to be ready to take control at all times; the China-market ADAS capabilities that Chinese brands advertise are not what UK-sold cars provide, and the underlying technology is further from mass-market reliability than the marketing implies.