Most Ethereum L2s Won't Survive 2026—Is Consolidation Around Base/Arbitrum/Optimism Good for Ethereum?

General
1

I just read the latest 21Shares forecast and honestly, it hit different than I expected. After spending 6 years working on L2 infrastructure, seeing data that says most Ethereum scaling networks won’t make it through 2026 feels like both vindication and tragedy.

The Numbers Don’t Lie

Right now, Base, Arbitrum, and Optimism are processing nearly 90% of all Layer 2 transactions. Base alone has captured over 60% of L2 activity. Meanwhile, Arbitrum is sitting on over $18 billion in TVL—more than any other L2. And here’s the kicker: Base was the only profitable L2 in 2025, earning around $55 million while most others were burning through funding just to keep the lights on.

We launched 50+ rollups over the past two years. After the incentive programs dried up, most of them became ghost towns. The users weren’t there for the tech—they were there for the yield farming opportunities.

Why These Three Are Winning

From a technical perspective, the winners aren’t accidents:

Base has the Coinbase advantage—seamless onboarding, actual retail users (not just degen farmers), and the marketing power of a publicly-traded company. Their OP Stack implementation is solid, and they’re iterating fast.

Arbitrum has the most battle-tested tech. The ArbOS Dia upgrade in January 2026 shows they’re still pushing forward on gas fee predictability and chain capacity. Plus, their $14M security audit fund signals serious long-term thinking.

Optimism cut DA costs by over 50% by moving to blobs. Their Superchain vision is ambitious—instead of one L2, they’re building an ecosystem of interconnected chains.

The Uncomfortable Question

Here’s what keeps me up at night: Is this consolidation good for Ethereum?

On one hand, I get the efficiency argument:

  • Liquidity concentration means better trading conditions
  • Clear winners = easier developer choices
  • Proven security > theoretical decentralization
  • Users don’t want to bridge between 47 different chains

But on the other hand, we’re watching something that looks suspiciously like centralization:

  • Base is controlled by Coinbase (CEX-backed L2)
  • Most L2s still have centralized sequencers and admin keys
  • Stage 1 decentralization is years away for even the winners
  • We went from “Ethereum will scale via modular rollups” to “Ethereum will scale via 3 corporate-backed platforms”

The Profitability Problem

Only Base made money in 2025. Think about what that means: if you’re not profitable and you don’t have deep VC pockets or a CEX parent company, you’re toast. The RaaS (Rollup-as-a-Service) providers made it trivially easy to launch a chain, but nobody solved the business model problem.

Launching an L2 became as easy as deploying a smart contract. We got the Cambrian explosion of chains, but no sustainable ecosystem.

What Happens Next?

The Ethereum Foundation’s latest L1-L2 update (March 23, 2026) was telling. They’re pushing for Stage 1 minimum requirements and talking about “differentiated services.” Translation: not all L2s will decentralize equally, and that might be okay if they serve different purposes.

Maybe we need specialized chains:

  • Gaming chains optimized for throughput (okay with some centralization)
  • DeFi chains optimized for security (Stage 2 decentralization priority)
  • Payment chains optimized for cost (aggressive data compression)

Or maybe we’re just rationalizing the fact that decentralization is hard and expensive, and the market is choosing efficiency over ideology.

My Take

After working at both Polygon and Optimism, I’ve seen this movie before. The early internet had hundreds of protocols. TCP/IP won. Not because it was technically perfect, but because it was good enough and everyone converged on it.

Maybe Base/Arbitrum/Optimism are the TCP/IP of Ethereum scaling. Maybe consolidation around 3-5 winners is the pragmatic path forward. Or maybe we’re giving up on the modular vision too early and centralization risk will bite us later.

I’m genuinely torn on this.

What do you all think? Are we watching healthy market consolidation or the slow death of Ethereum’s decentralization promise?

For those building on L2s—which one did you choose and why? For users—do you even care which L2 you’re on, or do you just want low fees and fast transactions?

2

Lisa, great breakdown. As someone who’s been contributing to core Ethereum infrastructure since 2016, I think the consolidation we’re seeing is natural evolution, not failure.

Security Comes From Maturity

Here’s the thing that doesn’t get enough attention: Base, Arbitrum, and Optimism have battle-tested codebases. They’ve survived the pressure tests that only real user activity can provide. Most of the failed L2s had weak security assumptions from day one—centralized sequencers with no decentralization roadmap, trusted operators, single-party upgrade keys.

When you’re securing billions of dollars, you can’t iterate fast and break things. You need formal verification, extensive auditing, and time-proven architecture. The three surviving L2s have that. Most of the others didn’t.

The Stage 1/2 Framework Changes Everything

The Ethereum Foundation’s push for Stage 1 and Stage 2 decentralization is going to accelerate this consolidation, not slow it down. Most L2s can’t hit Stage 1 requirements (fraud proof systems, decentralized sequencer networks, secure upgrade mechanisms) because they don’t have the engineering resources or financial runway.

Arbitrum’s $14M security audit fund? That’s what it takes to do this right. Most L2s were trying to survive on $2M seed rounds.

Historical Parallel

I keep thinking about the early internet. We had hundreds of competing protocols in the 1980s and early 1990s. TCP/IP won not because it was technically perfect—it wasn’t. It won because it was good enough, had institutional backing, and achieved critical mass.

That’s what we’re watching with Base/Arbitrum/Optimism. They’re not perfect. Coinbase controlling Base’s sequencer is problematic. But they’re good enough, and they have the resources to keep improving.

The Real Question

Should Ethereum Foundation mandate minimum decentralization standards for the “official L2” designation? Right now, anything that settles to Ethereum L1 can call itself an “Ethereum L2.” Maybe we need gatekeeping—only chains that hit Stage 1 get the branding.

That would be controversial (decentralization doesn’t pick winners!), but it might prevent the next wave of rug pulls and abandoned projects.

What’s your take on minimum standards vs. “let the market decide”?

3

As someone currently building a Web3 startup (pre-seed, trying to figure out which L2 to launch on), I’m going to say something that might be unpopular: this consolidation is actually GOOD for builders like me.

Users Don’t Care About Decentralization Theater

Here’s what I learned from customer interviews: normal people don’t care about L2 tech. They want:

  • Low fees (under $0.50 per transaction)
  • Fast confirmations (instant, not “eventually”)
  • Reliability (network doesn’t go down during high traffic)

When I tried explaining the difference between optimistic and zero-knowledge rollups to potential users, their eyes glazed over. But when I showed them Base’s Coinbase onboarding (literally two clicks from CEX to L2), they got it immediately.

The Fragmentation Was Killing Us

Before this consolidation, we were looking at 50+ L2 options. Each one had:

  • Different bridging infrastructure (some worked, some didn’t)
  • Fragmented liquidity (DEX prices varied wildly between chains)
  • Different developer tooling (some had decent docs, most didn’t)
  • Uncertain longevity (is this chain still going to exist in 6 months?)

That’s a nightmare for product development. Now we have three clear options with known tradeoffs. I can actually make a decision and move forward.

Profitability Matters

Only Base made money in 2025. Think about what that means from a business perspective: if your infrastructure provider isn’t profitable, they’re either living on VC money (temporary) or subsidizing with parent company funds (also temporary).

Base has Coinbase backing. Arbitrum has enough TVL to sustain fees. Optimism has the OP Stack ecosystem. Everyone else was burning runway.

But Here’s My Concern

Are we just rebuilding Coinbase on blockchain rails? If Base captures 80% of L2 activity and Coinbase controls the sequencer, what’s actually decentralized?

I got into Web3 because I believed in permissionless innovation. If I need Coinbase’s approval (directly or indirectly) to succeed, that’s just Web2 with extra steps and worse UX.

Right now, I’m leaning toward launching on Base for traction, then moving to Arbitrum if we need more decentralization credibility. But I’m not thrilled about that tradeoff.

What are other founders doing? Which L2 did you choose and why?

4

Okay, so I have a slightly different perspective as someone who actually built on three different L2s in 2025 (spoiler: two of them are now basically ghost towns).

My Learning Experience (aka Mistakes)

In early 2025, I was so excited about the “modular blockchain” vision. I thought: more L2s = more experimentation = better for developers. So we launched the same dApp on:

  1. Arbitrum (for the security-conscious DeFi users)
  2. A smaller L2 that offered us a grant (won’t name names, but they’re sub-1000 daily users now)
  3. Base (after we saw where actual users were going)

Guess what happened? The small L2’s grant dried up, their incentive program ended, and 95% of users left. Now we’re maintaining three separate deployments but only Base and Arbitrum have meaningful activity.

User Confusion Is REAL

You know what’s really hard? Explaining to non-technical users which L2 they’re on and how to bridge. I’ve watched people:

  • Send funds to the wrong chain (permanently lost or expensive to recover)
  • Pay $30 in gas to bridge $50 USDC (because they didn’t understand L1 vs L2)
  • Get frustrated and go back to Coinbase (can’t blame them)

This consolidation makes my job easier. Fewer chains = better wallet support, simpler onboarding, less confusing UX.

But I’m Worried About Monoculture

Here’s what keeps me up at night: if Base dominates everything, we’re just rebuilding Coinbase.

I moved from TradFi tech to Web3 because I believed in permissionless innovation. I wanted to build stuff that couldn’t be shut down or censored. If 80% of Ethereum activity runs through Coinbase’s infrastructure, what did we actually accomplish?

Right now, our project is on both Arbitrum (more decentralized, better tech) and Base (more users, better UX). That sucks. We’re maintaining two codebases, double the testing, double the support burden.

My Ideal World

Honestly, I think 3-5 specialized L2s is the sweet spot:

  • A gaming chain optimized for throughput (okay with centralized sequencer if TPS is high)
  • A DeFi chain optimized for security (Stage 2 decentralization, formal verification)
  • A payments chain optimized for cost (sub-$0.01 transactions)
  • Maybe one or two others for specific use cases

That’s better than 50 identical L2s competing on incentive programs. But it’s also better than one Coinbase-controlled chain dominating everything.

The Composability Question

How do we preserve composability across the surviving L2s? Right now, moving assets between Arbitrum and Base requires bridging (slow, expensive, risky). If we’re going to have multiple L2s, we need better cross-chain infrastructure.

What do y’all think? Anyone else maintaining multi-L2 deployments?

5

Wow, these responses are exactly the kind of nuanced discussion I was hoping for. Let me synthesize what I’m hearing and add some technical context.

Brian’s Security Argument

You’re absolutely right that maturity matters for security. The data backs this up: Arbitrum’s ArbOS Dia upgrade in January showed they’re still innovating on fundamentals (gas fee predictability, chain capacity, Passkeys for mobile auth). That’s not “maintaining legacy code”—that’s continuous improvement from a team with resources.

The Stage 1/2 framework is going to separate serious L2s from side projects. Most chains can’t afford the engineering work to implement fraud proof systems correctly, let alone decentralized sequencer networks.

Question for you: Do you think EF should mandate Stage 1 as a minimum for “official” L2 designation, or would that be too heavy-handed?

Steve’s Business Reality

Your point about profitability is something we don’t talk about enough in technical circles. Only Base made money in 2025. That’s not a small detail—that’s existential.

The RaaS providers (Rollup-as-a-Service) made launching chains trivial, but nobody solved the business model. Most L2s were:

  • Subsidizing transactions to attract users (burning runway)
  • Paying for DA costs that exceeded fee revenue
  • Competing on incentive programs (temporary growth)

Base won because Coinbase can subsidize losses in the short term and has a massive user funnel. Arbitrum won because TVL generates sustainable fees. Optimism won because the Superchain ecosystem shares costs across multiple chains.

Everyone else lost.

Emma’s Multi-Chain Pain

Your story about deploying to three L2s is exactly what I’m seeing from teams. The “modular blockchain” vision sounded great in theory: specialized chains for different use cases, composable ecosystem, permissionless experimentation.

In practice: fragmented liquidity, broken bridges, confused users, and developer teams stretched thin.

I think your ideal (3-5 specialized L2s) is where we’re heading. But we need to solve cross-L2 composability. Right now, bridging between Arbitrum and Base is:

  • Slow (7-day fraud proof windows for some bridges)
  • Expensive ($10-30 depending on L1 gas)
  • Risky (bridge exploits are still a thing)

Optimism’s Superchain and Arbitrum’s Orbit are trying to solve this with same-stack interoperability. But that still fragments into “OP Stack ecosystem” vs “Arbitrum ecosystem.”

Looking Forward: What Defines Success?

Here’s what I think we need to figure out as a community:

For L2s to be successful long-term, what metrics matter?

  • TVL? (Arbitrum leads with $18B)
  • Transaction count? (Base leads with 60%+ of activity)
  • Profitability? (Only Base in 2025)
  • Decentralization? (Nobody is Stage 2 yet)
  • Developer activity? (Declining across all chains)

Right now, we’re optimizing for different things and calling them all “success.”

The uncomfortable truth: Maybe consolidation around Base/Arbitrum/Optimism is the pragmatic path forward, even if it compromises the original vision. The alternative might be 50 L2s with minimal activity, worse UX, and no clear winners.

I’m genuinely curious: for those of you building or using L2s, what would make you switch chains? Is it lower fees? Better decentralization? Specific dApps you want to use? Liquidity?

Let’s figure this out together.