Base, the Ethereum layer 2 network incubated by Coinbase, closed out the 30 days ending September 7, 2026 as the largest rollup in crypto by total value locked. On-chain trackers put its TVL at roughly $14.42 billion, edging past Arbitrum One’s $12.6 billion and pushing combined value on the two leading networks past $27 billion. The number that turned heads wasn’t the TVL figure itself, though. It was what Base paid Ethereum’s base layer to process all of it: about $8,800 in data, proof, and state-update fees over the entire month, while handling 292 million user operations.
Run the math and the average daily bill to Ethereum mainnet comes out to roughly $290. For a network moving tens of billions of dollars in value and hundreds of millions of transactions, that’s a strikingly small line item. It’s also the clearest evidence yet of how thoroughly rollup economics have shifted since Ethereum’s 2024 fee overhaul, and it raises a question the industry has been circling for two years: what happens to Ethereum’s own revenue model once nearly all its economic activity moves one layer up, onto infrastructure that a single publicly traded company effectively controls.
Base’s $14.4 Billion Moment: What Just Happened
Total value locked measures the assets, loans, and liquidity sitting inside a network’s smart contracts. It’s an imperfect metric (it can be inflated by looped lending or double-counted collateral) but it’s still the industry’s default yardstick for size, and by that yardstick Base now sits ahead of every other Ethereum scaling network. The $14.42 billion figure, drawn from the 30-day snapshot ending September 7, 2026, represents sustained growth rather than a single day’s spike, which matters because TVL numbers can swing hard on volatile trading days.
Arbitrum One, built on Offchain Labs’ Nitro stack rather than Base’s OP Stack, held about $12.6 billion over the same window. That’s close enough that the ranking could flip on any given week, but the trend line for most of 2026 has favored Base, largely on the strength of consumer and DeFi activity funneled through Coinbase’s own retail and institutional user base. Together the two networks account for more than $27 billion, a level of concentration that means whatever happens to Base or Arbitrum now moves the entire layer 2 sector, not just one project.
Breaking Down the Numbers: 292 Million Operations, $8,800 in Fees
The 292 million figure covers “user operations,” a broader category than simple transfers. It includes swaps, contract calls, account abstraction batches, and other interactions that Base’s infrastructure bundles and settles back to Ethereum in compressed form. That’s the entire point of a rollup: execute cheaply off-chain, then post a compact proof of what happened to Ethereum, which acts as the settlement and security layer underneath.
What Base pays for that settlement is what generated the headline. Roughly $8,800 across 30 days works out to a fraction of a cent per operation once spread across 292 million interactions. Before anyone assumes Ethereum is getting shortchanged, it’s worth remembering this is by design. Ethereum’s own roadmap has spent three years explicitly optimizing for rollups to pay less, on the theory that cheap L2 activity expands the addressable market for the whole ecosystem rather than shrinking Ethereum’s slice of a fixed pie.
| Metric | Figure | Context |
|---|---|---|
| Total value locked (Base) | $14.42 billion | Largest of any Ethereum L2 as of the Sept. 7, 2026 snapshot |
| Total value locked (Arbitrum One) | $12.6 billion | Second-largest L2, built on Nitro rather than OP Stack |
| Combined Base + Arbitrum TVL | >$27 billion | The bulk of tracked L2 value sits on just these two networks |
| Base 30-day user operations | 292 million | Transactions plus contract calls and batched interactions |
| Base’s 30-day L1 fees paid | ~$8,800 | Data, proof, and state-update costs settled on Ethereum mainnet |
| Implied average daily L1 cost | ~$290/day | Derived from the 30-day total fee figure |
| Rollup framework | OP Stack | Optimistic rollup design shared with Optimism’s OP Mainnet |
Why Ethereum Collects So Little From Its Biggest Rollup
The short answer is a single upgrade: Dencun, which activated on Ethereum mainnet on March 13, 2024, and shipped EIP-4844, commonly known as “blobs” or proto-danksharding. Before Dencun, rollups posted their transaction data as regular calldata, competing for space and gas with every other Ethereum transaction. CoinMarketCap’s academy explainer notes that L2 data posting accounted for roughly 10% of Ethereum’s total fee revenue in the period leading up to the upgrade, a meaningful chunk of the network’s economic activity.
Blobs gave rollups a dedicated, much cheaper data lane that expires after roughly 18 days rather than living on-chain forever. According to Trust Wallet’s analysis of the blob rollout, the change cut layer 2 transaction costs by a factor of 100 to 200 almost immediately, with fees on some networks falling from around $1 to under a penny. Base’s $8,800 monthly fee bill on 292 million operations is the practical, two-years-later result of that shift: a network can now process a small nation’s worth of daily transactions while paying Ethereum less than a mid-size household spends on groceries in a month.
The OP Stack Behind Base’s Architecture
Base runs on the OP Stack, the open-source rollup framework originally built by the team behind Optimism and now maintained as shared public infrastructure through the Optimism Collective. Coinbase’s own developer documentation describes Base as “an Ethereum Layer 2, incubated by Coinbase and built on the open-source OP Stack, that allows developers to easily deploy reliable, secure applications on a scaling solution with low transaction fees,” a description that has held up well against the September numbers.
Sharing a codebase with Optimism and other OP Stack chains isn’t just a technical convenience. It’s a bet on what’s sometimes called the “Superchain” thesis: that a family of interoperable rollups running compatible software will out-compete isolated, bespoke chains by pooling security research, tooling, and liquidity bridges. Base’s TVL lead is now the strongest evidence that bet paid off for at least one member of that family, even as Arbitrum’s separately built Nitro stack shows the OP Stack doesn’t have a monopoly on scale.
From Testnet to $14 Billion: Base’s Three-Year Climb
Base’s rise wasn’t instant. Coinbase announced the network’s testnet on February 23, 2023, framing it in its own blog post as a way to offer “a secure, low-cost, developer-friendly way for anyone, anywhere, to build decentralized apps or ‘dapps’ onchain.” Developer access opened on July 13, 2023, and the public mainnet launched on August 9, 2023, timed to a month-long “Onchain Summer” marketing push. At launch, TechCrunch’s coverage described Coinbase as the second-largest crypto exchange by trading volume at the time, a detail that mattered because it meant Base launched with a built-in retail funnel most rollups never get.
That funnel shows up in the current numbers. Coinbase has spent three years routing its own consumer products, from wallet features to onchain trading tools, through Base rather than treating it as a side experiment. A support article from Eco, an independent crypto payments firm, sums up the relationship plainly: “Base is Coinbase’s Ethereum L2, built on the OP Stack.” Three years after testnet, that simple description now applies to a $14.4 billion network.
| Date | Milestone | Effect |
|---|---|---|
| Feb. 23, 2023 | Base testnet announced by Coinbase | Introduced an OP Stack-based L2 aimed at onchain app builders |
| Aug. 9, 2023 | Base public mainnet launch (“Onchain Summer”) | Opened the network to all developers and retail users |
| Mar. 13, 2024 | Dencun upgrade activates EIP-4844 (“blobs”) | Cut L2 data-posting costs by roughly 100x-200x industry-wide |
| 2024-2025 | L2 total value locked consolidates | Base and Arbitrum One separate from the rest of the field |
| Sept. 7, 2026 | Base TVL reaches $14.42B on ~$8,800 in 30-day L1 fees | Base becomes the largest Ethereum L2 by locked value |
Base vs. Arbitrum vs. Optimism vs. zkSync: How They Actually Compare
Base and Arbitrum One now sit well clear of the rest of the Ethereum layer 2 field on locked value, but the two aren’t technical twins. Base runs the OP Stack, an Optimistic rollup design that assumes transactions are valid unless someone proves otherwise during a challenge window. Arbitrum One uses Offchain Labs’ own Nitro stack, also Optimistic but with a different fraud-proof implementation and its own separate token and governance structure through the Arbitrum DAO.
Optimism’s own flagship chain, OP Mainnet, and zero-knowledge rollups such as zkSync Era trail both leaders by locked value, though exact rankings shift week to week on trackers like L2Beat. The gap matters less for raw TVL bragging rights than for what it says about strategy: ZK rollups bet on cryptographic proofs that can theoretically reach a higher security tier faster, while Optimistic designs like Base and Arbitrum bet on speed to market and ecosystem gravity, arguing users care more about available liquidity and apps than about which proof system sits underneath.
Base’s specific edge over its OP Stack sibling, Optimism’s own chain, comes down almost entirely to distribution. Coinbase has millions of existing account holders it can nudge onto Base with a tap inside an app they already use daily. Arbitrum and Optimism have had to build that user base organically through grants, incentive programs, and independent app ecosystems. Base’s TVL lead is, in that sense, less a story about superior rollup engineering and more a story about what happens when a regulated exchange with an enormous retail base decides to become infrastructure.
Market Impact: What This Means for Coinbase and Ethereum
For Coinbase, a publicly traded company, Base’s TVL milestone is a business metric as much as a technical one. Base generates revenue for Coinbase through sequencer fees, the small charges users pay for their transactions to be ordered and batched before settlement on Ethereum. As Base’s operation count climbs into the hundreds of millions per month, that fee stream, however small per transaction, becomes a growing and largely proprietary line of business sitting on top of public, shared infrastructure.
For Ethereum, the picture is more complicated. The $8,800 Base paid over 30 days is trivial next to the $14.4 billion it’s helping secure, and that gap is exactly the tension critics have flagged since Dencun shipped. Ethereum validators and ETH holders capture value primarily through gas fees and issuance dynamics tied to base-layer activity. If the bulk of economic activity keeps migrating to L2s that pay almost nothing back, Ethereum’s own fee revenue could stay structurally thin even as the ecosystem it anchors keeps growing, a dynamic some researchers have already started calling the rollup “value capture” problem. It’s a very different kind of risk from the exploits that dominate most crypto security headlines, where stolen private keys now outpace code bugs as the top attack vector, but it’s arguably a bigger long-term threat to the broader cryptocurrency ecosystem’s economics.
The Centralization Question: A Public Company Running Core Infrastructure
Base’s sequencer, the component that orders and batches transactions before they’re posted to Ethereum, is currently operated by Coinbase. That’s standard for most rollups at this stage of the industry’s development, but it means a single, US-listed, regulated company controls the ordering of transactions across a network now holding more value than most mid-size national payment systems. If Coinbase’s sequencer goes down, is compelled by a regulator, or simply prioritizes its own transactions, Base users have limited recourse beyond an exit through Ethereum’s underlying security guarantees.
That risk hasn’t stopped adoption, and it may never meaningfully slow it, since most users treat rollups the way they treat cloud providers: as infrastructure to trust rather than infrastructure to audit personally. But it’s a structural feature worth flagging now, while Base is merely the largest L2, rather than after it becomes something closer to a systemic dependency for onchain finance.
Security Stage Ratings and the Decentralization Roadmap
L2Beat, the independent analytics site that tracks rollup security assumptions alongside TVL, classifies networks into stages based on how much a chain still depends on a centralized operator versus trustless, on-chain fraud or validity proofs. Most major rollups, Base included, still sit in early stages of that framework rather than the fully decentralized “Stage 2” tier, which requires permissionless proof submission and strict limits on how fast an operator can unilaterally change the rules.
Coinbase has publicly committed, alongside the broader OP Stack development community, to progressively decentralizing Base’s sequencer and proof systems over time. Whether that timeline accelerates now that Base carries the “largest L2” title, and the added scrutiny that comes with it, is one of the more concrete things to watch over the next several quarters rather than the next several years.
Cost Per Operation: A Rough Reality Check
Base, 30-day window ending Sept. 7, 2026
User operations: 292,000,000
L1 fees paid to Ethereum: $8,800
Implied cost per op: $8,800 / 292,000,000 ~= $0.00003
Implied daily L1 cost: $8,800 / 30 ~= $293/dayThat back-of-envelope math is the clearest way to see why builders keep choosing rollups over deploying directly on Ethereum mainnet. It also explains why competition among L2s has shifted almost entirely away from fees, which are now near-zero across the leading networks, and toward security guarantees, available liquidity, and how tightly a chain is woven into an existing user base. Readers who want a hands-on look at how those fee savings translate into practice can walk through our own layer 2 scaling setup guide, or compare the mechanics against Arbitrum’s own onboarding path in our Arbitrum bridge setup walkthrough.
What Comes Next: Five Predictions for Layer 2 Scaling
- TVL concentration keeps rising. With Base and Arbitrum already holding more than $27 billion between them, expect the gap between the top two networks and the rest of the field to widen rather than narrow over the next year, as liquidity and app developers chase existing user density.
- Decentralization pressure grows with scale. As Base’s numbers draw more mainstream financial press, expect louder calls, including from within the OP Stack community itself, for Coinbase to accelerate sequencer decentralization on a fixed public timeline rather than an open-ended roadmap.
- Ethereum’s fee-capture debate intensifies. Expect renewed proposals from Ethereum researchers aimed at letting the base layer capture more value from the L2 activity it secures, since the current $8,800-on-$14-billion ratio is not a sustainable long-term revenue model for validators.
- Smaller L2s consolidate or shut down. Networks without a built-in distribution advantage, the way Base has Coinbase, will face growing pressure to merge liquidity, share infrastructure, or wind down, as the fee-driven differentiation that used to separate rollups keeps eroding.
- Regulatory attention shifts up a layer. As exchanges like Coinbase run more of the infrastructure their users’ assets depend on, expect regulators to start asking questions about L2 sequencers as critical financial infrastructure, not just about the exchanges that operate them.
What Developers and Traders Should Watch
For builders, Base’s numbers are less an argument for choosing one specific chain and more a confirmation that rollup economics have crossed a threshold where fees no longer decide anything on their own. Anyone still comparing L2s primarily on gas costs, the way the industry did in 2023, is optimizing for a variable that’s already been mostly solved. The more useful questions now are which network’s liquidity actually matches an app’s target users, and which one is closer to shedding its training wheels on decentralization.
For traders and institutions, the TVL race is a useful proxy for where onchain activity is settling, but it’s worth remembering TVL can move fast in both directions. A single large lending unwind or a security incident on either network could shift the rankings within a single news cycle. Base itself isn’t immune to that risk: DeFi protocols built on top of it have already been targeted, including a price-oracle exploit on a Base-based lending protocol that drained $8.7 million earlier this year, a reminder that a rollup’s own security doesn’t automatically extend to every app built on top of it.
Frequently Asked Questions
What is Base, in plain terms?
Base is an Ethereum layer 2 network, or rollup, incubated by Coinbase and built on the open-source OP Stack. It processes transactions off Ethereum’s main chain and then posts compressed proof of that activity back to Ethereum, which handles final settlement and security.
Why does Base’s TVL matter?
Total value locked is the industry’s standard measure of how much capital, in loans, liquidity, and deposits, sits inside a network’s smart contracts. At $14.42 billion, Base now holds more than any other Ethereum L2, which affects everything from developer interest to how much liquidity is available for trading and lending.
How can Base process 292 million operations while paying so little to Ethereum?
Ethereum’s March 2024 Dencun upgrade introduced dedicated, low-cost “blob” data space for rollups, cutting L2 data-posting costs by roughly 100x-200x industry-wide. Base compresses hundreds of millions of user interactions into that cheap data channel rather than posting each one individually.
Is Base decentralized?
Not fully. Base’s sequencer, which orders transactions before they’re settled on Ethereum, is currently operated by Coinbase. Base and the wider OP Stack community have committed to progressive decentralization, but most rollups, Base included, remain in early stages of security-decentralization frameworks like L2Beat’s.
How is Base different from Arbitrum One?
Both are Optimistic rollups, but they run different software stacks: Base uses the OP Stack, while Arbitrum One uses Offchain Labs’ Nitro stack. Arbitrum also has its own token and DAO governance structure, while Base does not currently have a standalone token.
Does Base have its own cryptocurrency token?
No. Base does not have a native token separate from ETH. Transaction fees on Base are paid in ETH, and Coinbase, rather than a token-holder DAO, currently operates the network’s sequencer.
What happens to Ethereum’s revenue if L2 fees stay this low?
It’s an open debate. Ethereum’s base-layer revenue from securing rollup activity has stayed structurally small since Dencun, even as the value it secures keeps growing, which has fueled ongoing research into new fee or value-capture mechanisms for the base layer.
What’s the difference between an Optimistic rollup and a zero-knowledge rollup?
Optimistic rollups, like Base and Arbitrum, assume transactions are valid and allow a challenge window for anyone to dispute them with fraud proofs. Zero-knowledge rollups, like zkSync Era, generate cryptographic proofs upfront that mathematically verify transaction validity, which can allow faster finality but has historically been more complex to build.
