The L2 Wars Are Over and Arbitrum Already Won

For the last three years, the Ethereum L2 landscape felt like a horse race. Arbitrum vs. Optimism vs. zkSync vs. StarkNet vs. Polygon zkEVM. Everyone picked a horse. VCs poured billions into competing rollup teams. The "L2 wars" were the hottest narrative in crypto infrastructure.

I'm going to say something controversial: the wars are over. Arbitrum won.

And the only thing that might challenge it isn't another L2 at all. It's a chain built by Coinbase.

The TVL Tells the Story

DeFi Llama data doesn't lie. Arbitrum consistently sits as the largest L2 by total value locked. It's not close. Arbitrum hosts the deepest DeFi ecosystem outside of Ethereum mainnet, with major deployments from Aave ($1.1 billion on Arbitrum alone), GMX, Uniswap, and dozens of native protocols.

Optimism is a distant second. zkSync Era launched with massive hype and a token airdrop that brought millions of users. But the post-airdrop activity collapsed. StarkNet has impressive technology but a fraction of the usage. Polygon's various zkEVM efforts have struggled to gain traction.

The pattern is clear. Users and developers gravitate toward the chain with the most liquidity and the most protocols. Arbitrum had first-mover advantage in the optimistic rollup space, captured the DeFi ecosystem early, and has been compounding that lead ever since.

Why Arbitrum Won

It comes down to three things: timing, ecosystem, and pragmatism.

Timing. Arbitrum launched on mainnet in August 2021, catching the tail end of DeFi summer. Developers who were frustrated with Ethereum's gas costs had a production-ready L2 to migrate to. Optimism was live but less mature. ZK rollups were still in testnet. Arbitrum captured the first wave of serious DeFi migration.

Ecosystem. GMX chose Arbitrum. That single decision brought perpetual futures traders, which brought liquidity, which brought more protocols. Treasure DAO brought gaming. Camelot brought a native DEX. The ecosystem developed its own identity rather than being a "cheap Ethereum copy." When you have unique applications that only exist on your chain, users have a reason to bridge in and stay.

Pragmatism. Arbitrum chose the optimistic rollup model, which was simpler and more battle-tested than ZK approaches. It shipped when others were still building. Offchain Labs (the team behind Arbitrum) prioritized "works now" over "theoretically superior." That pragmatism paid off massively.

The ARB Token Problem

Here's where the story gets complicated. Arbitrum the chain is winning. ARB the token? Not so much.

ARB has underperformed BTC and ETH since its airdrop. The token's primary utility is governance, and governance tokens have been a terrible investment across the board. There's no fee-sharing mechanism. No burn. No strong reason to hold ARB other than voting on proposals.

The Arbitrum DAO controls a massive treasury, but using it effectively has been a challenge. Governance proposals have funded some good things (incentive programs, grants) and some questionable things. The typical DAO problems apply. Voter apathy, whale domination, and misaligned incentives.

If you're bullish on Arbitrum the ecosystem, buying ARB isn't necessarily the best way to express that view. You might be better off using protocols on Arbitrum (earning yields on Aave, trading on GMX) than holding the governance token.

The Base Wild Card

Now for the uncomfortable part. While Arbitrum was winning the L2 wars among crypto-native teams, Coinbase was quietly building Base. And Base is eating.

Base launched in August 2023 and has grown at an astonishing rate. It benefits from something no other L2 has: direct integration with the largest US crypto exchange. When Coinbase users bridge to L2, they bridge to Base. When Coinbase introduces new features, they build on Base first.

The numbers are hard to ignore. Base's daily transaction count regularly exceeds Arbitrum's. The user base is different, more retail, more on-ramp focused, and less DeFi-heavy. But users are users, and transaction volume is transaction volume.

Base doesn't have a token (yet?). That's both a strength and a weakness. No token means no governance overhead, no speculative distraction, and no regulatory headache. But it also means the "L2 wars" aren't being scored the way most people expected.

Where Optimism Fits

Optimism has taken a different approach entirely. Instead of competing head-to-head with Arbitrum for DeFi TVL, it's building the Superchain, a network of interconnected L2s built on the OP Stack.

Base is built on the OP Stack. Worldcoin's chain is built on the OP Stack. Several other projects have adopted it. Optimism's strategy is to be the "framework" rather than the "chain," earning sequencer revenue from the broader ecosystem rather than competing on any single metric.

It's clever. The OP token accrues value from all Superchain chains, not just Optimism mainnet. But it's also a bet that framework value exceeds application value, which is far from proven.

The ZK Rollup Disappointment

I need to address the elephant in the room. ZK rollups were supposed to win the L2 wars.

The technology is genuinely better. Validity proofs provide instant finality instead of the 7-day challenge period of optimistic rollups. They're more trustless, more scalable in theory, and more elegant.

But theory doesn't ship products. zkSync Era and StarkNet both launched with impressive technology and disappointing adoption. The developer experience was harder. The tooling was less mature. The EVM compatibility was partial or absent (in StarkNet's case).

After their token airdrops, both saw massive drops in activity. The farmers left. The developers who needed battle-tested infrastructure stayed on Arbitrum.

ZK rollups might eventually win on technical merit. But "eventually" is starting to feel like a cope. Arbitrum's lead is widening, not narrowing.

My Take

Arbitrum won the L2 wars by being good enough, fast enough. Not the most innovative technology. Not the best token model. Not the most hyped marketing. Just a reliable, deep, well-functioning DeFi ecosystem that developers and users could count on.

The real competition going forward isn't between crypto-native L2s. It's between Arbitrum and Base. Between a decentralized, community-governed L2 and a corporate-backed, Coinbase-integrated one. That's a much more interesting fight, and it'll play out over the next two to three years.

If you're building on an L2, Arbitrum is still the safe choice. If you want retail distribution, Base is compelling. Everything else is increasingly niche.

The L2 landscape won't be winner-take-all. But it'll be winner-take-most. And right now, Arbitrum is that winner.