Every few years, a new battery technology arrives with the promise of rendering everything before it obsolete. The 4680 large-format cylindrical cell — unveiled by Tesla at its 2020 Battery Day with headline claims of 5× the energy, 6× the power, and 16% more range — is the latest contender to spark this debate. Headlines since have oscillated between breathless enthusiasm ("the end of 18650") and sobering reality checks ("4680 underperforms supplier cells").
So which is it? Is the 4680 genuinely the death knell for 18650 and 21700 formats, or just another tool in the cylindrical battery toolbox — brilliant for some applications, irrelevant for others?
The answer, as with most things in battery engineering, lies not in the hype but in the physics, the economics, and the iron grip of installed infrastructure. Let's walk through exactly what the 4680 changes — and what it doesn't.
The Cylindrical Evolution: Three Generations, Three Philosophies
To understand where 4680 fits, you need to understand the two generations that came before it.
18650: The Pioneer (1990s). First commercialized by Sony, the 18650 (18mm × 65mm) became the world's standard cylindrical lithium-ion cell. It powered the laptop revolution, then was adapted by Tesla for the original Roadster and Model S (which used roughly 7,000 cells per pack). Today, 18650 cells deliver 1500mAh to 3600mAh, with high-drain variants capable of 25A to 35A continuous discharge. The technology is mature, the supply chain is globally distributed, and the cost per cell is the lowest of any cylindrical format.
21700: The Optimizer (2017). Jointly promoted by Tesla and Panasonic for the Model 3, the 21700 (21mm × 70mm) was designed to strike the optimal balance between energy density, power, and manufacturing cost. Compared to the 18650, the 21700 offers approximately 35% more capacity under the same material conditions and roughly 20% higher volumetric energy density. With typical capacities of 3000mAh to 5800mAh and high-drain ratings reaching 30A to 45A, the 21700 has become the preferred format for premium power tools, high-performance flashlights, portable power stations, and EV applications.
4680: The System-Level Play (2020–present). The 4680 (46mm × 80mm) is categorically different. Where the 18650→21700 transition was evolutionary, the 4680 represents an attempt at a system-level revolution. A single 4680 cell delivers roughly 85–98Wh of energy — about 5× the capacity of a 21700 cell. A 100kWh EV battery pack using 21700 cells requires approximately 4,400–5,000 individual cells; the same pack built with 4680 cells needs only 960–1,000 cells. Fewer cells mean fewer welded connections, fewer points of failure, simpler battery management, and dramatically faster pack assembly.
But the 4680 isn't just bigger — it's architecturally different. The defining innovations are the tabless electrode design and the dry electrode manufacturing process.
What Makes 4680 Fundamentally Different: Tabless Design and Dry Electrode
In a conventional cylindrical cell like the 18650 or 21700, a thin metal tab connects the electrode jellyroll to the cell terminal. At high currents, that tab becomes a bottleneck — current funnels through a tiny cross-sectional area, generating concentrated resistance and localized heating. This is why traditional 21700 cells have internal resistance around 30mΩ with a single tab design.
The 4680's tabless design transforms the entire edge of the electrode into a current collector. Instead of one narrow tab, current flows through multiple paths along the full circumference of the electrode. Internal resistance drops to below 15mΩ — roughly half that of a conventional 21700. This means less heat generation, more uniform thermal distribution, and dramatically improved fast-charging capability, enabling 5–15 minute ultra-fast charging in theory.
The dry electrode (DBE) process, acquired by Tesla from Maxwell Technologies, eliminates the solvent coating and drying steps required in traditional wet electrode manufacturing. By skipping the energy-intensive drying ovens — which can account for a significant portion of a gigafactory's footprint and energy consumption — the DBE process promises lower manufacturing costs, reduced factory footprint, and minimized environmental impact.
These two technologies, combined with the sheer size of the cell, are the foundation of Tesla's original vision: a battery that costs half as much to produce while delivering 5× the energy and 6× the power.
The Reality Gap: What the Data Actually Shows
Tesla's Battery Day presentation promised the moon. Five years later, the measured performance tells a more complicated story.
Energy density: worse, not better. Tesla's 4680 cells produced at Giga Austin have a nominal gravimetric energy density of approximately 244 Wh/kg. The Panasonic 2170 cells the 4680 was designed to replace sit at 269 Wh/kg — meaning the first-generation 4680 is roughly 13% worse on the metric that matters most for EV range. While Tesla claims newer versions improve on this, independent testing of production cells has not yet confirmed those claims.
Real-world range impact. When Tesla began swapping LG-supplied 2170 packs for its own 4680 packs in European Model Y vehicles, the direct consequence was a WLTP range drop from 661 km to 609 km — a 52 km reduction, or roughly an 8% downgrade — in the same vehicle with the same aerodynamics and motors.
Charging performance: the biggest disappointment. The tabless design was supposed to enable ultra-fast charging. Real-world data from first-generation 4680 Model Y vehicles (2023) showed the opposite: heat buildup caused charging power to drop below 100 kW after reaching just 35% state of charge, with a 10–80% charging time exceeding 40 minutes — dramatically worse than the 2170-equipped Model Y Long Range, which completes the same charge in roughly 27–30 minutes.
Production struggles. Tesla's internal 4680 program has been plagued by manufacturing difficulties. Musk himself admitted at the 2025 shareholder meeting that the dry electrode process turned out to be "way harder" than expected. A key supply partner effectively wrote down its 4680 contract by 99% in late 2025, signaling a dramatic pullback in Tesla's production ramp-up plans. Meanwhile, Tesla has shifted 4680 procurement to external suppliers — LG Energy Solution began pilot production at its Ochang plant in South Korea, and Panasonic launched mass production at its Wakayama factory in Japan.
The implication is clear: the 4680 is not yet delivering on its founding promises, and the transition is proving far slower and more difficult than the original 2020 timeline suggested.
The Consumer Cylindrical Cell Market: Why 18650 and 21700 Aren't Going Anywhere
When battery enthusiasts ask whether 18650 will become obsolete, they're usually thinking about their flashlights, power tools, power banks, and DIY projects. The answer is unambiguous: 18650 and 21700 are not merely surviving — they're growing, and the growth is driven by precisely the applications the 4680 cannot serve.
Consider the data: the global 18650 and 21700 lithium battery market was valued at approximately 86.185 billion RMB (roughly 12billionUSD)in2024,andisprojectedtoreach∗∗210.305billionRMB∗∗(roughly12billionUSD)in2024,andisprojectedtoreach∗∗210.305billionRMB∗∗(roughly29 billion USD) by 2030 — growing at a compound annual rate of over 15%. The broader cylindrical lithium-ion battery market is expanding from roughly 10.8billion∗∗in2024towardanestimated∗∗10.8billion∗∗in2024towardanestimated∗∗64.1 billion by 2034, growing at roughly 19.5% CAGR.
These aren't the market dynamics of a format heading for obsolescence. They're the dynamics of a format diversifying into more applications than ever.
Why? Because the laws of physics that make 4680 compelling for EVs make it a non-starter for most consumer applications. The 4680 cell is 46mm in diameter and 80mm in length — it's the size of a small beverage can. It doesn't fit in a flashlight. It doesn't fit in a power tool battery pack. It doesn't fit in a laptop. It doesn't fit in a portable power bank designed to slide into a backpack pocket. The 4680 is designed for vehicle-scale integration — structural battery packs, cell-to-chassis architectures, and thermal management systems built around large-format cells.
Meanwhile, the applications driving 18650 and 21700 demand continue to expand:
Power tools need cells with high continuous discharge ratings (25A–35A) in 18650 and 21700 formats that fit existing pack housings
Consumer electronics — laptops, portable speakers, vacuum cleaners — rely on 18650 cells in standardized, space-constrained compartments
Flashlights and portable lighting demand cells ranging from compact 14500/18650 formats to high-capacity 21700 options
Portable power stations and USB-C power banks increasingly adopt 21700 cells for their superior energy-to-size ratio
E-bikes, e-scooters, and light electric vehicles use 18650 and 21700 cells in standardized pack configurations
Medical devices and industrial equipment require the reliability and safety track record that only mature, proven cell formats can provide
Across these applications, the 18650 and 21700 market segments cover consumer electronics, drones, power tools, new energy vehicles, and more. None of these devices can physically accommodate a 46mm-diameter cell.
Today, our 18650 Battery collection and 21700 Battery collection serve customers building everything from single-cell flashlights to multi-cell e-bike packs — applications where the 4680's size makes it physically impossible to use, regardless of how its performance improves.
The Infrastructure Lock-In: Why Format Transitions Take Decades, Not Years
Beyond physics, there's an economic force that strongly favors 18650 and 21700 retention: installed infrastructure lock-in.
Hundreds of millions of devices in the field today are designed around the 18650 and 21700 form factors. Every cordless drill, every vape mod, every tactical flashlight, every laptop battery pack, every portable power station — each one has a battery compartment machined to a specific diameter and length. Changing the cell format means redesigning the device, retooling the production line, requalifying the supply chain, and managing the transition for existing customers with legacy devices.
These transitions happen, but they happen on decade-long timescales. The 18650 was commercialized in the 1990s and remains the most widely produced cylindrical lithium-ion format in 2025. The 21700 was introduced in 2017 and is only now reaching widespread adoption in consumer devices beyond EVs. The 4680, even under the most optimistic adoption scenarios, will not meaningfully penetrate consumer applications for at least another decade — and likely far longer.
For users choosing between high-capacity and high-drain cells for their specific devices, the practical decision remains firmly rooted in 18650 and 21700 formats. Our High-Drain 18650 collection and High-Capacity 18650 collection organize cells by their real-world CDR and runtime characteristics precisely because these are the specifications that matter for the devices people actually use today. If you're unsure which category your device falls into, our guide on high-drain vs high-capacity batteries: power cells vs energy cells selection guide provides a detailed current-based selection framework.
Coexistence, Not Replacement: The Three-Format Future
The most realistic projection for the cylindrical cell market over the next 10–15 years is not replacement — it's stratification by application.
For electric vehicles and grid-scale storage, 4680 and other 46-series large-format cells represent the future. The system-level advantages — fewer cells per pack, simpler BMS, structural integration, and dramatically faster pack assembly — are transformative for these applications. Industry projections suggest that by 2025, 4680 battery installations could reach approximately 264GWh, potentially capturing about 22% of the global power battery market and 68% of the cylindrical battery market by installed capacity. The period from 2025 to 2030 is expected to mark an acceleration phase in 4680 industrialization as production lines gradually come online.
For consumer electronics, power tools, portable power, and DIY applications, 18650 and 21700 will remain dominant for the foreseeable future — not because they're superior in absolute terms, but because they're the right size for the job. The continued growth of the 18650/21700 market at 15%+ CAGR through 2030, in parallel with the 4680 ramp, demonstrates that these are complementary markets, not competing ones. The vast ecosystem of devices designed around 18mm and 21mm diameters, combined with the massive installed base of chargers, BMS solutions, pack assembly tools, and community knowledge, creates a powerful lock-in that no new format can overcome quickly.
For builders interested in larger-format cells for higher-voltage pack designs — where 4680 is still largely inaccessible to the consumer market — LiFePO₄ options in 32140 and 32700 formats offer an excellent intermediate step. Our IFR32140 LiFePO₄ battery collection and IFR32700 LiFePO₄ battery collection feature large-format cells with industry-leading cycle life and intrinsic safety, ideal for solar storage, RV power, and marine applications. For a deeper understanding of why LiFePO₄ chemistry offers inherent safety advantages that no BMS can replicate, read our detailed guide on the olivine structure of LiFePO₄ and why it won't catch fire.
For the consumer cylindrical cell market specifically, the most likely outcome is a layer cake: 4680 at the top for large-format, high-energy applications (primarily automotive), 21700 in the middle for premium portable power and high-drain tools, and 18650 at the base for the broadest range of consumer devices — mature, cost-optimized, and irreplaceable for millions of existing products.
What This Means for You as a Battery Buyer
If you're shopping for cells today for a flashlight, power bank, power tool pack, or DIY project, the 4680 is irrelevant. Not because it isn't interesting — but because it won't physically fit in your device, and even if it could, the cell availability for consumers is essentially zero.
Your practical decision remains firmly in the 18650 and 21700 domain. The questions that matter are the same as they've always been:
What is your device's maximum continuous current draw?
Do you need high-drain or high-capacity cells?
Are you building a pack that requires matched cells with specific CDR and IR characteristics?
What chemistry is appropriate for your safety requirements — NMC or LiFePO₄?
For pack builders, understanding series-parallel configurations is essential to answering these questions. Our guide on series-parallel combinations: how to calculate total voltage and capacity for 3S2P, 4S3P packs provides the calculation framework for any pack design. For power tool rebuilds specifically — where ultra-high-drain cell selection is non-negotiable — our guide on cell selection for power tool battery packs: why ultra-high drain is essential walks through the exact CDR requirements for different tool types.
For cells, chargers, and accessories across all form factors — from 14500 to 21700, and including the large-format LiFePO₄ options that serve as the consumer-accessible bridge to bigger cells — visit our OneAndes complete store.
The Bottom Line
The 4680 is a genuinely important innovation in cylindrical cell technology — but it's an automotive and grid-storage innovation, not a consumer electronics one. Its tabless design, dry electrode manufacturing, and system-level integration advantages will reshape how EV battery packs are designed and assembled in the coming decade. But its sheer physical size (46mm × 80mm) precludes it from the vast ecosystem of devices that depend on 18650 and 21700 cells.
The 18650 is not dying. The 21700 is accelerating. The 4680 is growing into its own, separate market.
The cylindrical battery market of 2035 will almost certainly contain all three formats — and likely more — each optimized for different points on the spectrum from compact portability to maximum single-cell energy. The 4680 will not obsolete the 21700 any more than the 21700 obsoleted the 18650. It will simply add another tool to the toolbox — brilliant for the jobs it was designed for, and irrelevant for the ones it wasn't.
