Price Comparison of the Best Lithium Battery Manufacturers
BloombergNEF's annual battery price survey released on December 9, 2025 gave a global average of $108/kWh. This number is widely cited, but it mixes China's $84/kWh, Europe's $131/kWh, and North America's $121/kWh together. In actual procurement, no one pays this price. Three markets operate with three completely independent pricing systems.
CATL's Cost Black Box
CATL quoted ¥0.32/Wh for 314Ah energy storage cells to CNPC Jichai in Q3 2025. After media like Battery China reported this, many analysts believed it was a loss-leader to capture market share. Using industry-standard cost models, the materials plus manufacturing cost for 314Ah LFP cells comes to roughly ¥0.28-0.30/Wh. A ¥0.32/Wh quote implies a gross margin of only 6-12%, far below normal levels for the battery industry.
CATL's Q3 report shattered this assumption. Company filings showed net profit attributable to shareholders of 49.034 billion yuan for the first three quarters, up 36.2% year-over-year, equivalent to roughly 180 million yuan per day. The power battery business gross margin was 23.94%. These two figures together are puzzling. If ¥0.32/Wh were the standard price, it couldn't support a 23.94% gross margin. The quote to CNPC Jichai might be strategic low pricing to lock in a major customer, not representative of overall pricing. Or CATL's actual manufacturing cost might be much lower than industry estimates, since external analysts work from public information and industry averages. Another possibility is that energy storage and power batteries have different profit structures, with energy storage for volume and power batteries for profit, averaging out to acceptable margins overall. Which explanation is correct, outsiders can't see clearly. CATL doesn't publish cost breakdowns by product line. What we can directly observe is only the result: according to SNE Research November 2025 data, CATL's global power battery installation share from January to October was 37.9%. Share expanding, profits growing, prices declining. Three things happening simultaneously.
The Naxtra sodium-ion battery is a noteworthy development in 2025. Mass production began in June, with 175Wh/kg energy density and cycle life exceeding 10,000 times. Large-scale deployment will occur in 2026 across battery swap stations, passenger vehicles, commercial vehicles, and energy storage. The significance of sodium-ion isn't performance. It's the completely different cost structure. Sodium's crustal abundance is over 1,000 times that of lithium, no cobalt is needed, and aluminum foil can replace copper foil. Material costs are theoretically 30-50% lower than LFP. If Naxtra can achieve hundred-GWh-scale production in 2026, it will establish a new floor below lithium-ion prices. By then, the prices we're discussing today might become historical highs.
Where does CATL's pricing power come from? Scale is part of it: as the world's largest battery manufacturer, fixed cost amortization effects are significant. Vertical integration is part of it: from lithium mines to cathode materials to cell manufacturing, many links are done in-house or through deep partnerships. But these only explain part of it. The rest, outsiders can't see. We can only infer from results: CATL can profit at prices that would cause competitors to lose money.
There's a question rarely discussed: is CATL's cost advantage temporary or permanent? If it comes from scale effects, competitors could catch up by expanding to similar scale. If it comes from accumulated technology and management, the gap will widen. Looking at data from the past five years, the answer leans toward the latter. CATL's share rose from around 25% in 2020 to around 38% in 2025, while profit margins didn't decline with intensified competition. They remained stable or even increased. Competitors expanded capacity but couldn't close the gap.
The battery industry has a concept called the "learning curve," where costs decline by a certain percentage every time output doubles. This rule has been verified in solar and memory chips. But CATL's situation seems unexplainable by learning curve alone.
Learning curve implies that followers can catch up just by scaling up, but actually the cost curves of chasers and CATL's are declining in parallel. The gap isn't converging. This suggests CATL's advantage isn't just "doing more" but also "doing better," and "doing better" is harder to replicate than "doing more."
The 587Ah large cell is another example. Deliveries began in June 2025, mainly for large-scale energy storage projects. Larger cells mean lower system integration costs but also higher manufacturing difficulty. That CATL can achieve acceptable yields is itself a demonstration of capability.
Outside observers often discuss what CATL's "moat" is. Some say scale, some say patents, some say customer relationships. But perhaps the most important moat is something less glamorous: continuous manufacturing process improvement. Every production line, every process step, every quality control node has room for optimization. Each optimization looks small individually, but accumulated they become a huge cost gap. And this optimization is tacit. It can't be written into patents, can't be copied, can only be figured out bit by bit on your own. CATL has been doing this for over a decade.
BYD
BYD's battery prices can't be directly compared to CATL's because BYD isn't a battery company. It's a car company. Blade Battery 2.0 was released in April 2025 with 190-210Wh/kg energy density, 800V high-voltage platform support, and 15% cost reduction versus the first generation. But technical specs aren't the key point. The key is that BYD sold 4.27 million vehicles in 2025, and every Blade Battery-equipped car is internally absorbed capacity. When CATL expands capacity, it has to worry about customers being poached. When BYD expands capacity, the customer is itself. Tesla and BYD signed a 10GWh supply agreement. This news itself says something: Tesla spent over a decade on battery self-development, then decided to buy from BYD.
BYD's model is hard to replicate. It requires being top-tier simultaneously in both automotive and batteries, demanding staggering resource investment and organizational capability. Toyota has money but can't do batteries well. LG makes decent batteries but doesn't make cars. Companies that can do both things well? There might be only BYD in the whole world.
EVE Energy and the Logic of Large Cells
Most industry attention focuses on chemistry, discussing LFP versus ternary, sodium-ion versus solid-state. Few people pay attention to cell format. This is a blind spot. Chemistry determines energy density and cycle life, but cell format determines system integration costs. For energy storage applications that don't require high energy density but are extremely cost-sensitive, format innovation might be more valuable than chemistry innovation.
EVE Energy validated this judgment with the 628Ah large cell. China Power News reported in December 2024 that EVE Energy's 60GWh super energy storage factory in Jingmen was commissioned, and the 628Ah "Mr. Big" large cell began mass production. The Jingmen factory was purpose-built for large-format cells. Production lines, logistics, and testing equipment are all optimized for the 628Ah specification. The cost advantage of large cells isn't in the cell itself but at the system level. A 100MWh energy storage station using 280Ah cells needs about 360,000 cells, corresponding to 360,000 welding points, 360,000 monitoring nodes, and complex thermal management systems. Using 628Ah cells, the count drops to about 160,000 cells. Welding workload, assembly time, quality testing costs, and thermal management complexity all decrease accordingly. EVE estimates the format brings a 10-15% system-level cost advantage. Battery China reported in December 2025 that cumulative shipments of 628Ah cells exceeded 300,000 units. In September 2025, the world's first energy storage station using 628Ah cells was connected to the grid in Lingshou, Hebei. Market validation came faster than expected.
The numbers from the Malaysia factory deserve more attention. EVE expects the Malaysia base can achieve cell costs of ¥0.13-0.14/Wh, while domestic equivalent products cost about ¥0.22/Wh. A factory outside China, with costs 39% lower than China.
This data is worth pausing to think about. Western tariffs on Chinese batteries follow this logic: Chinese batteries are cheap because of subsidies, cheap labor, and concentrated supply chains. Tariff Chinese products and manufacturing will be forced to return. But EVE's Malaysia factory shows that Chinese battery companies can replicate or even optimize China's cost structure in Southeast Asia. Malaysian labor is cheaper than China's, electricity is cheaper than China's, and there are local government investment incentives. If Chinese companies build factories in Malaysia, the batteries produced aren't counted as Chinese-made, and costs are even lower than domestic. So who exactly are the tariffs protecting? CATL is building factories in Germany and Hungary, Gotion is building in the US, SVOLT is building in Thailand. Chinese battery companies' global deployment is faster than Western governments can tighten trade barriers.
There's an implicit assumption behind tariff policies: manufacturing capability follows factories. Where the factory is, that's where the capability is. But reality is that manufacturing capability follows people and organizations. Chinese companies bring management teams, process know-how, and supplier relationships to Malaysia, and Malaysian factories can replicate the efficiency of Chinese factories. Tariffs can restrict product flow but not capability flow.
The 628Ah specification is now adopted by multiple Chinese manufacturers and is becoming the new standard for energy storage. Once a standard is established, followers have to follow. Gotion started mass production deliveries of Unified Cells to Volkswagen in November 2025, and CALB won orders from Hyundai and Volkswagen. When Volkswagen and Hyundai made these procurement decisions, their governments were discussing supply chain autonomy and reducing dependence on China. Procurement departments look at cost data; policy departments discuss a different logic.
Korean Manufacturers
The 2025 story of Korea's three battery manufacturers can be summarized through LG's experience. LG Energy Solution revenue was 23.67 trillion won, down 7.6% year-over-year. Operating profit actually rose to 1.35 trillion won, up 133.9% year-over-year. But this profit growth wasn't because business improved. It mainly came from IRA tax credits and layoff savings. In December, Ford first terminated a $6.5 billion battery supply contract, citing cancellation of electric F-150 Lightning production. Nine days later, the FBPS contract was also gone: $2.7 billion. In less than a month, LG lost contracts worth more than half of 2024 sales. Ford's cancellation reason was straightforward: the math didn't work. Ford knew what Chinese batteries cost, knew what LG cost, compared the two, and saw that electric pickup trucks would lose money no matter what.
LG's current capacity utilization is only around 51%. Half the capacity sits idle, fixed costs spread over less output, unit costs pushed up, competitiveness further declining. Chinese leaders run at around 90% utilization.
Samsung SDI lost 1.5 trillion won on battery business in the first three quarters. SK's BlueOval joint venture fell apart in December. It was an $11.4 billion project. Ford took the Kentucky factory, SK kept Tennessee. Electrive reported that the combined share of Korea's three dropped from 19.6% to 15.8%.
The capacity utilization gap creates a vicious cycle. Low utilization means high unit costs, weak competitiveness, orders lost, utilization even lower. To break this cycle requires either dramatically expanding sales or dramatically shrinking capacity. Expanding sales needs orders, and the Ford situation shows major customers can leave anytime. Shrinking capacity means layoffs and asset write-downs, which is hard to execute for public companies.
One view is that Korean manufacturers can survive through technology differentiation: high-nickel ternary, 4680 large cylindrical, solid-state batteries. But Chinese manufacturers are also working on these areas, and progress isn't slow. The premise of technology differentiation is being able to do what others can't, or doing it many years earlier. If Chinese manufacturers are only two or three years behind, then so-called technology differentiation is just a time gap, not a moat. LG's 46-series cylindrical cells supposedly have a 300GWh order pipeline, but China's BAK, Lishen, and EVE are all making cylindrical cells at lower prices.
Panasonic's situation is slightly better. The Kansas factory started production in July 2025. It was a $4 billion investment with 32GWh capacity, combined with Nevada for 73GWh total. Tesla is the biggest customer, but Tesla is also increasing purchases from CATL and BYD. Full production timing slipped 6-12 months from the original March 2027 plan. After Trump canceled the 30D tax credit, the market isn't as good. Toyota's solid-state battery might be a variable. They planned for 10GWh mass production at Japanese factories in 2026, with specs of 450-500Wh/kg, 1,200km range, and 10-minute charge to 80%. Toyota has over 8,000 solid-state battery patents, has missed timelines before, and the technology is genuinely difficult.
Energy Storage
Energy storage battery prices can't be summarized in a single number. In 2025, this market split into three pieces with completely different pricing logic.
The Chinese domestic market is brutal. 314Ah cell average prices dropped to around ¥0.29/Wh in the first half, with some quotes reaching ¥0.26/Wh. Low prices result from overcapacity. Chinese energy storage cell capacity is about 557GWh, global installation demand about 250GWh, capacity is more than double demand. Second and third-tier manufacturers have few choices: either ship at a loss, keep lines running, and bet competitors die first; or just shut down. In the second half, lithium carbonate rebounded from 70,000/ton to 84,000/ton, overseas demand pulled inventory down, and GGII predicts energy storage cells will rise 5-10% in 2026.
The US is another world. BloombergNEF data shows energy storage system costs of $219-236/kWh, more than three times China. Policy is the main reason. The July 2025 OBBBA Act tightened domestic content requirements. Projects using Chinese batteries must start construction before end of 2025 to avoid restrictions. Policy is artificially creating a price floor. European residential storage system average prices dropped from €1,332/kWh in early 2023 to €711/kWh in H2 2025. Chinese brands like Huawei, Growatt, and Pylontech dominate the market. Northvolt went bankrupt in March 2025, with $5.8 billion debt, having previously raised over $12 billion. This was Sweden's largest industrial bankruptcy in history.
Three markets, three price sets. China $66-73/kWh, US $219-236/kWh, a threefold gap. For US energy storage developers, the same 100MWh project costs roughly $6.6-7.3 million with Chinese batteries, $21.9-23.6 million with domestic batteries, a gap of around $15 million. Some developers choose to break ground before end of 2025, using Chinese batteries under more relaxed policies. Some choose to wait, betting policy will loosen. Others choose US domestic batteries, passing high costs through to electricity prices.
How long can this price split persist? Short-term, no convergence is visible. Chinese manufacturers have no incentive to raise prices, US policy shows no sign of loosening, European domestic capacity has basically exited. Three markets will continue operating on their own logic until some external shock breaks the equilibrium. It might be large-scale commissioning of Chinese manufacturers' overseas capacity, might be US policy reversal, might be some new technology making existing capacity obsolete. But these are all "might," not "will."
Procurement
Chinese leading LFP cell manufacturers quote in the ¥0.28-0.35/Wh range, translating to $40-50/kWh. Battery packs are around $84/kWh. Korean manufacturers are 30-50% higher for the same specs, still 25-40% more expensive after accounting for IRA credits. A car with a 60kWh battery choosing Chinese versus Korean suppliers means a cost difference of roughly $1,500-2,800. Multiply by hundreds of thousands in annual sales, and this number determines whether an automaker can survive in the EV market. Energy storage gaps are even larger. China $66-73/kWh, US $219-236/kWh, threefold.
SNE Research data shows Chinese manufacturers held 68.9% of global power battery installation share from January to October 2025. Korea's three combined dropped to 15.8%. BloombergNEF predicts global battery pack average price will drop to $105/kWh in 2026; Goldman Sachs thinks $80-82/kWh is achievable. $80/kWh is the tipping point where EVs and ICE vehicles reach total cost of ownership parity. But these are all global averages. The gap between Chinese and non-Chinese manufacturers won't shrink as average prices fall.
The manufacturing capability gap remains. Subsidies can alleviate symptoms, tariffs can delay the inevitable, but neither solves the fundamental problem. To change this situation, Western manufacturers need not more subsidies or higher tariffs, but genuine catch-up in manufacturing capability. This requires time, capital, and strategic resolve. Most manufacturers lack all three.
The competitive landscape of the battery industry may already be set. Not that followers have absolutely no chance, but the window for catching up is closing. The advantages CATL and BYD have built in cost, scale, and technology are mutually reinforcing: cost advantage brings market share, market share brings scale effects, scale effects support R&D investment, R&D investment solidifies technology leadership, technology leadership converts back to cost advantage. Once this positive feedback loop forms, breaking it from outside is extremely difficult. Unless there's disruptive technological change, like solid-state batteries actually maturing, the current landscape will persist for quite a long time.