Over 50 OP Stack Chains, Dozens of Arbitrum Orbits, and Only 5-10 Rollups With Real Users - The RaaS Industry Created a Chain Glut and Liquidity Fragmentation Is the Price We Pay

[crypto_chris]

The Numbers Don’t Lie: We Built Too Many Chains

I’ve been tracking L2 metrics obsessively for the past eighteen months, and the picture that emerges is both predictable and damning. The Rollup-as-a-Service (RaaS) industry — led by Conduit, Caldera, Gelato, and AltLayer — made it trivially easy to launch a new chain. Under 30 minutes, minimal capital, full-stack deployment. And the market responded exactly as you’d expect when you remove friction from supply creation without addressing demand: we got a glut.

Let me lay out the data.

The OP Stack Superchain alone now hosts over 50 chains. Base, Zora, Mode, Mint, Cyber, Redstone, Lisk, Metal, Swan, and dozens more. On the Arbitrum Orbit side, we have dozens of additional chains — Xai, Degen Chain, Sanko, Proof of Play, and a growing tail of project-specific rollups. Polygon CDK spawned Immutable zkEVM, Astar zkEVM, and others. The total number of distinct L2s and L3s in production today is staggering.

But here’s the kicker: the top 5 L2s — Arbitrum, Base, Optimism, Blast, and Scroll — capture over 80% of all L2 TVL and transaction volume. That’s not a healthy long tail. That’s a power law with a wasteland at the bottom.

Most RaaS-deployed chains have under $10M in TVL and fewer than 1,000 daily active users. Some have single-digit DAUs on any given day. These aren’t ecosystems — they’re ghost towns with block explorers.

Liquidity Fragmentation Is the Structural Risk

This isn’t just an aesthetic problem or a “too many chains” meme. Liquidity fragmentation is a structural risk for the next cycle, and I don’t think enough people are taking it seriously.

Here’s why it matters:

1. Tokens on multiple rollups fragment liquidity, complicating trading and lending. If a token exists on Arbitrum, Base, Optimism, Blast, and three app-specific rollups, the liquidity for that token is split seven ways. Market makers need to maintain positions on each chain. Arbitrageurs need to bridge constantly. Users on thin-liquidity chains get worse execution.

2. Even small trades can significantly impact prices on thin-liquidity chains. I’ve watched $5,000 swaps move prices 3-5% on some of these smaller L2 DEXs. That’s not a functional market — that’s a price manipulation surface. When you have a DEX with $200K in TVL on a chain with 300 daily active users, every trade is a market-moving event.

3. DeFi composability breaks across chain boundaries. The magic of DeFi on mainnet Ethereum was composability — flash loans touching Aave, Uniswap, and Curve in a single transaction. That atomic composability doesn’t exist across rollups. Every cross-chain interaction requires bridging, waiting, and trusting intermediaries.

4. Developer attention fragments alongside liquidity. Protocol teams now need to decide which 5-10 chains to deploy on, maintain separate deployments, manage cross-chain governance, and handle bridge-related edge cases. This isn’t scaling — it’s multiplication of operational burden.

The Solutions Being Proposed

The industry isn’t blind to this problem. Several approaches are being developed:

Intent-based trading is probably the most promising near-term solution. Instead of users manually bridging and swapping, they express an intent (“I want to swap X for Y at this price”) and solvers compete to fill it across chains. UniswapX, Across Protocol, and the broader solver network are pushing this forward. But intent-based systems introduce their own centralization risks — the solver market tends toward oligopoly.

Concentrated liquidity mechanisms (Uniswap V3/V4 style) help individual pools operate more efficiently with less capital, but they don’t solve the cross-chain fragmentation problem. You still need the liquidity to be on the right chain at the right time.

Cross-chain routing algorithms from aggregators like Li.Fi and Socket try to find optimal paths across chains, but they add latency, bridge risk, and gas overhead. They’re patches, not cures.

The UAW20 universal abstract token standard has been proposed to address fragmentation at the token level — creating a single logical token that exists across multiple rollups with unified liquidity. It’s conceptually elegant but requires broad adoption and standardization that the fragmented ecosystem may not be able to coordinate on.

The Uncomfortable Question

Here’s what I keep coming back to: did we need 50+ OP Stack chains? Did we need dozens of Arbitrum Orbits? The RaaS thesis was that every application would want its own chain for sovereignty, customization, and dedicated blockspace. But the reality is that most applications don’t generate enough activity to justify a dedicated chain, and the ones that do (gaming, social) could often be served by a shared chain with dedicated lanes.

The RaaS industry created supply-side abundance without demand-side growth. We made chains cheap to launch but didn’t make users cheap to acquire. The result is a landscape where 90% of rollups are solutions in search of problems, burning through treasury while their TVL languishes in the low millions.

I’m not saying rollups are bad. I’m saying the market needs consolidation, and the chains that survive will be the ones that can attract and retain genuine liquidity and users — not the ones that got the cheapest RaaS deal.

The next cycle will be defined by which chains can break through the fragmentation trap. The rest will quietly sunset, and the RaaS providers will pivot to “chain migration services” for the survivors.

What’s your take — is the chain glut a temporary growing pain or a structural failure of the modular thesis?

[blockchain_brian]

The Modular Thesis Isn’t Wrong — The Market Just Hasn’t Caught Up Yet

Chris, your data is solid but your framing is doing a lot of heavy lifting. Let me offer a technical rebuttal that I think recontextualizes the “chain glut” narrative.

First, let’s separate the modular thesis from the RaaS execution. The modular blockchain thesis — that separating execution, data availability, consensus, and settlement into specialized layers produces better systems — is architecturally sound. What happened with RaaS is that the tooling got ahead of the demand curve. That’s a timing mismatch, not a thesis failure.

The internet went through an identical phase. In the late 1990s, anyone could register a domain and stand up a website in minutes. Millions did. Most of those websites had zero traffic. Nobody concluded that “the internet created a website glut.” The infrastructure maturation curve has a messy middle phase, and we’re in it.

Second, the 80% concentration metric needs context. Yes, the top 5 L2s capture 80%+ of TVL and transactions. But that’s also true of every nascent technology market. The top 5 cloud providers capture 80%+ of cloud revenue. The top 5 mobile apps capture 80%+ of screen time. Power law distributions are the natural state of network-effect markets, not evidence of failure.

What matters is whether the long tail creates value that couldn’t exist otherwise. And here, I think the answer is clearly yes:

• Immutable zkEVM exists specifically because gaming transactions have different requirements (high throughput, low cost, NFT-optimized) than DeFi transactions. Putting Illuvium and Gods Unchained on mainnet Arbitrum would be suboptimal for both gaming users and DeFi users.
• dYdX Chain moved to its own Cosmos chain because orderbook-based perpetual trading needs sub-second finality and custom gas economics that shared chains can’t provide.
• Ronin (Axie Infinity) processes millions of gaming transactions that would congest any shared chain and benefits from custom gas token economics.

These are legitimate use cases for dedicated chains. The problem is that for every Immutable, there are 20 chains that launched because RaaS made it free, not because the application demanded it.

Third, on the technical solutions — you’re underselling the progress. The fragmentation problem is real, but the infrastructure to address it is being built in parallel:

Shared sequencing through projects like Espresso Systems creates a common ordering layer across rollups. If multiple rollups use the same sequencer, cross-rollup atomic transactions become possible. The Superchain’s shared sequencer vision, if executed, would let all OP Stack chains share liquidity natively. This doesn’t exist yet at scale, but the architecture is being actively developed.

Interop protocols are maturing. The OP Stack’s native interop messaging system, combined with Superchain USDC (where Circle mints natively on each chain with instant cross-chain transfers), represents a different model than the “bridge everything” approach. When tokens are natively minted across chains with coordinated supply, you don’t have fragmentation — you have distribution.

ERC-7802 and cross-chain token standards are being developed specifically to create unified token representations across rollups. The UAT20 / ERC-7802 approach lets a token maintain a single logical supply across chains with seamless transfers. This is early but architecturally clean.

Fourth, and this is the part where I partially agree with you: the fragmentation problem is currently severe and the solutions are not yet deployed at scale. The gap between “this architecture can solve fragmentation” and “fragmentation is actually solved” is measured in years, not months.

Where I push back is on the implication that this is a permanent structural failure. The modular stack is going through what I’d call the “pre-interop phase” — we’ve built the execution layers but haven’t shipped the communication layers. It’s like having 50 islands with no bridges or ferries. The islands aren’t the problem; the missing transportation infrastructure is.

The real risk isn’t too many chains — it’s that interop arrives too late. If we spend another 18 months with fragmented liquidity while the market enters a bull cycle, users will consolidate onto 3-4 chains organically (likely Arbitrum, Base, and one or two others), and the interop infrastructure will arrive for an ecosystem that’s already calcified around a few winners. The window for the “unified multi-chain” vision is finite.

My prediction: we’ll see meaningful consolidation in 2026-2027, not because chains fail technologically, but because liquidity gravity pulls everything toward 4-6 major chains. The remaining chains will either specialize (gaming, social, enterprise) or quietly deprecate. And the interop stack (shared sequencing + native cross-chain tokens + intent-based routing) will mature just in time to make the surviving multi-chain ecosystem actually functional.

The modular thesis survives, but with far fewer chains than the RaaS providers projected. It’s a pruning, not a funeral.

[layer2_lisa]

As Someone Building L2 Infrastructure, Here’s What Both of You Are Missing

I work on L2 infrastructure daily, and while I agree with a lot of Chris’s data points, I think the “chain glut” framing misunderstands what’s actually happening at the engineering level. Brian’s closer to the mark, but let me add some nuance from the builder’s perspective.

The 50+ OP Stack chains number is misleading without context. Not all chains are created equal, and lumping Base (billions in TVL, millions of users) with a testnet-grade chain that launched last Tuesday distorts the picture. A more useful breakdown:

• Tier 1 (3-5 chains): Billions in TVL, millions of monthly active users, full DeFi ecosystems. Base, Arbitrum One, OP Mainnet, maybe Blast and Scroll.
• Tier 2 (10-15 chains): Tens to hundreds of millions in TVL, legitimate use cases, growing but niche. Mode, Mantle, Linea, zkSync Era, Manta, Zora.
• Tier 3 (30+ chains): Under $10M TVL, application-specific or experimental. Many of these are intentionally small — they’re app-chains for specific games, social platforms, or enterprise use cases. They don’t need billions in TVL.

The Tier 3 chains having low TVL isn’t a failure state if they were never designed to be general-purpose DeFi chains. Xai exists for gaming. Degen Chain exists for a specific community. Redstone exists for on-chain gaming infrastructure. Measuring them by DeFi TVL is like measuring a microservice by the amount of traffic it handles compared to the monolith — it misses the point of decomposition.

Now, where I agree the problem is real: general-purpose L2s with no differentiation.

If you launched an OP Stack chain in 2024 with the pitch “we’re a general-purpose L2 with low fees” and no specific application thesis, you’re in trouble. The market doesn’t need chain #47 that does the same thing as Arbitrum but with 1/1000th the liquidity. Those chains represent the actual chain glut, and yes, most of them will die.

On the fragmentation solutions — I want to be pragmatic about timelines.

I’ve been working with the OP Stack interop specifications, and I can tell you the engineering is further along than most people realize but also further from production than the optimistic timelines suggest.

The Superchain interop model works like this: chains in the Superchain share a message-passing protocol that enables cross-chain calls without traditional bridges. Combined with ERC-7802 (the CrosschainERC20 standard), you can have a token that’s natively available across all Superchain chains with atomic transfers. No wrapping, no bridges, no liquidity fragmentation for that token.

The catch: this only works within the Superchain ecosystem. Arbitrum Orbit chains, zkSync chains, and Polygon CDK chains don’t benefit from OP Stack interop. So you solve fragmentation within one ecosystem while the cross-ecosystem fragmentation persists.

This is where intent-based systems become critical. Chris mentioned that solvers tend toward oligopoly, and he’s not wrong — but the solver market is more nuanced than a simple “centralization bad” take:

• Competitive solver markets (like UniswapX’s Dutch auction design) create price competition even with a small number of solvers. You don’t need 1,000 solvers — you need 5-10 competitive ones.
• Solver specialization is emerging: some solvers optimize for speed, others for price, others for specific chain pairs. This is healthy market structure.
• The user experience improvement is massive. Even with imperfect solver decentralization, a user who clicks “swap” and gets their tokens on the destination chain in 5 seconds without thinking about bridges is a 100x UX improvement over the current state.

On the UAT20 / universal abstract token standard: I’ve reviewed the proposals, and I think this is the right long-term direction but it requires coordination that’s politically difficult. Every chain needs to adopt the standard, every token issuer needs to upgrade, and the cross-chain mint/burn mechanics need to be secured. That’s a multi-year migration at best.

My pragmatic outlook for builders:

1. If you’re building a new L2 today, you need a specific application thesis. “General purpose L2” is no longer a viable pitch. Your chain needs to serve a specific vertical (gaming, social, DeFi, enterprise) with specific technical advantages (custom precompiles, modified gas economics, specialized data availability).

2. If you’re a DeFi protocol, focus your liquidity on 3-5 chains maximum. Deploy on Arbitrum, Base, and your best-performing chain. Don’t spread thin across 15 rollups hoping intent-based systems will fix the UX later.

3. If you’re an RaaS provider, start building the abstraction layer now. The value in 2027 won’t be “deploy a chain in 30 minutes” — it’ll be “deploy a chain that’s natively interoperable with the ecosystem from day one.”

The ecosystem is maturing, and maturation always involves attrition. But framing it as a “structural failure” overstates the case. We overbuilt, the market will prune, and the surviving infrastructure will be better for it. That’s how technology cycles work.

The fragmentation is painful now but solvable. The question is whether we solve it before users give up and consolidate onto Solana and Base by default.

[startup_steve]

This Is a Classic Overcapacity Cycle — And the Playbook Is Predictable

I’ve been through enough technology market cycles to recognize what’s happening here, and the RaaS chain glut fits a pattern that’s repeated across every infrastructure market I’ve seen.

The overcapacity playbook goes like this:

1. New infrastructure technology dramatically reduces the cost of deployment (RaaS makes launching a chain nearly free)
2. Supply floods the market because the barrier to entry collapsed faster than demand could grow
3. A brutal shakeout follows where 80-90% of entrants die or get acquired
4. The survivors consolidate and the market matures around an oligopoly of 5-7 major players
5. The infrastructure providers pivot from “deploy new” to “migrate and manage existing”

We’ve seen this in cloud computing (hundreds of cloud providers in 2010, now it’s AWS/Azure/GCP/Alibaba), in SaaS (thousands of CRM tools in 2012, now it’s Salesforce/HubSpot/a few verticals), and in mobile apps (millions of apps launched, but the average phone uses 30 apps and the top 10 capture 80% of engagement).

The RaaS market is in stage 2, heading into stage 3. The question isn’t whether consolidation happens — it’s how fast and who survives.

Let me frame this from a business perspective that the technical folks in this thread might be missing:

The real cost of a chain isn’t deployment — it’s ecosystem development. Conduit can deploy your OP Stack chain in 30 minutes for a few hundred dollars a month. But building an ecosystem around that chain — attracting developers, bootstrapping liquidity, acquiring users, establishing partnerships, maintaining infrastructure, funding grants — costs tens of millions of dollars and years of sustained effort.

Base succeeded not because Coinbase used special technology, but because Coinbase had 110+ million verified users, a trusted brand, deep regulatory relationships, and the ability to onboard users directly from its exchange. The OP Stack deployment was table stakes. The distribution advantage was everything.

Most RaaS-deployed chains have none of these distribution advantages. They launched a chain and then discovered that “build it and they will come” doesn’t work when there are 50 other chains saying the same thing. The resulting ghost towns aren’t a technology problem — they’re a go-to-market problem.

The market dynamics of chain consolidation are already visible:

Liquidity gravitates toward liquidity. This is the single most powerful force in crypto markets, and it works against fragmentation. When a trader needs to execute a $1M swap, they go where the pool depth supports it without 5% slippage. That’s Arbitrum or Base, not chain #47 with $200K in DEX TVL. Every day that passes without cross-chain liquidity solutions, the liquidity moat around the top chains deepens.

Developer attention follows user attention. Protocols deploy where users are. Users go where protocols are. This creates a flywheel that’s nearly impossible for small chains to break into. I’ve talked to protocol teams who maintain deployments on 10+ chains, and they universally say that 90% of their volume comes from 2-3 chains. The long-tail deployments exist for “optionality” but generate negligible revenue.

VC funding for “generic L2” is drying up. I sit on the advisory board of a crypto fund, and I can tell you that in 2024, “we’re launching an L2” was a fundable pitch. In 2026, it needs to be “we’re launching an L2 for [specific vertical] with [specific distribution advantage] and [specific technical differentiation].” The generic L2 pitch is dead for fundraising.

What does the consolidation look like in practice?

I expect the following over the next 12-18 months:

1. Quiet sunsets. Most of the sub-$10M TVL chains won’t announce they’re shutting down. They’ll just stop updating their infrastructure, let their RaaS contracts lapse, and the chain will become a ghost. Some might continue running with zero users indefinitely because the cost is so low.

2. Ecosystem mergers. Some chains will “merge” by migrating their users and liquidity to a larger chain in exchange for ecosystem grants or governance tokens. This is already happening informally — projects that launched on small L2s are redeploying on Base or Arbitrum.

3. RaaS provider pivots. Conduit, Caldera, and Gelato are smart companies. They’ll pivot from “launch your chain” to “manage your multi-chain deployment” and “migrate your chain to a more sustainable configuration.” The real money in RaaS will be B2B infrastructure management, not chain launches.

4. Vertical chain winners. The chains that survive in the long tail will be the ones with genuine vertical specialization. Immutable for gaming, maybe a chain for social (Farcaster’s architecture is interesting here), potentially enterprise chains for regulated institutions. These chains succeed because their applications require custom chain-level features, not because they wanted sovereignty for sovereignty’s sake.

The fragmentation problem solves itself through consolidation. When the market naturally prunes from 100+ chains to 10-15 relevant ones, the liquidity fragmentation becomes manageable. Intent-based trading across 10 chains is a tractable engineering problem. Intent-based trading across 100+ chains is a coordination nightmare.

Chris is right that we have a chain glut. Brian is right that the modular thesis isn’t dead. Lisa is right that the engineering solutions exist. But the market dynamics that will actually resolve this are simpler and more brutal than any of them suggest: most chains will fail, liquidity will consolidate, and the survivors will benefit from less competition for users and capital.

That’s not a bug — it’s how markets work.

[defi_diana]

Let Me Tell You What Fragmentation Actually Looks Like From Inside DeFi

Everyone in this thread is talking about fragmentation abstractly — TVL numbers, chain counts, theoretical solutions. I want to ground this in what it actually means for DeFi protocols, liquidity providers, and users who are trying to trade, lend, and farm across this fractured landscape. Because from where I sit, the damage is already here and it’s worse than the metrics suggest.

The DEX Depth Problem Is Existential for Small Chains

Chris mentioned that $5,000 swaps can move prices 3-5% on thin-liquidity chains. Let me make that more concrete.

I’ve been monitoring DEX pools across 15+ L2s for a yield optimization strategy. On a typical small OP Stack chain with $500K in total DEX TVL spread across 5-10 pools, the largest ETH/USDC pool might have $100K in liquidity. In a concentrated liquidity pool (Uniswap V3 style), that $100K might have $30K concentrated in the active price range.

At that depth, a $10,000 swap — not even a large trade — creates 2-3% price impact. A $50,000 swap creates 8-12% slippage. This isn’t a market. It’s a puddle that splashes when anyone steps in it.

For any sophisticated DeFi user, these chains are unusable for anything beyond micro-transactions. And the circular problem is obvious: nobody provides liquidity because there are no traders, and nobody trades because there’s no liquidity. The chain needs an external catalyst (incentives, partnerships, a killer app) to bootstrap, and most RaaS-deployed chains don’t have one.

Lending Markets Suffer Even More Than DEXs

The fragmentation impact on lending is particularly brutal, and I don’t think enough people discuss this.

A healthy lending market requires:

• Sufficient collateral diversity and depth
• Reliable oracle price feeds (which require DEX liquidity for on-chain pricing)
• Enough borrowing demand to generate meaningful yields for depositors
• Liquidation infrastructure (bots, MEV searchers) to maintain solvency

On a chain with $5M in TVL and 500 daily active users, none of these conditions are met. The oracle feeds are unreliable because there isn’t enough on-chain trading to generate accurate prices. The liquidation infrastructure is thin because there isn’t enough liquidation volume to justify running bots. And the yield for depositors is near-zero because borrowing demand is negligible.

I’ve seen lending protocols deploy on small L2s with initial incentive programs, attract $2-3M in deposits, watch borrowing demand flatline at $200K, and quietly wind down. The economics simply don’t work when the user base can’t support a functioning lending market.

Liquidity Mining Becomes a Race to the Bottom

Here’s where fragmentation creates a particularly insidious dynamic for DeFi yields.

When 50+ chains are all competing for the same limited pool of DeFi users and capital, each chain needs to offer incentives to attract liquidity. This creates an “incentive arms race” where:

1. Chain A offers 50% APY in native token rewards to attract LPs
2. Chain B sees liquidity flowing to Chain A and offers 75% APY
3. Mercenary capital rotates between chains, chasing the highest incentive
4. Neither chain builds sticky liquidity because the capital leaves when incentives drop
5. Both chains burn through treasury faster than they can build organic demand

This dynamic is directly caused by fragmentation. If all that liquidity were concentrated on 3-5 chains, protocols wouldn’t need to outbid each other for the same capital. The incentive spending would be lower, the liquidity would be stickier, and the yields would be more sustainable.

The yield farmers know this. The sophisticated capital allocators I interact with have a simple heuristic: farm the incentives on new chains, extract the rewards, bridge back to Arbitrum or Base. They treat small chains as temporary yield extraction opportunities, not ecosystems to invest in. This is rational behavior in a fragmented market, but it’s destructive to the chains that are trying to bootstrap.

Cross-Chain DeFi Composability Is a Fantasy Right Now

Brian mentioned that cross-chain composability is being built. Let me be blunt about where we actually stand:

• Flash loans are single-chain only. The atomic composability that enabled the most capital-efficient DeFi strategies on Ethereum mainnet simply doesn’t exist cross-chain.
• Liquidation cascades don’t propagate across chains. If a collateral asset crashes on Arbitrum, the lending positions on Base using that same asset might not get liquidated promptly because the oracle updates and liquidation bots are chain-specific.
• Yield strategies that span multiple protocols need to be on the same chain. A strategy that deposits into Aave, borrows against it, and LPs the borrowed assets in Uniswap needs all three to be on the same chain for gas efficiency and atomic execution.

Intent-based systems help with simple swaps, but they don’t solve DeFi composability. They solve the “I want token X on chain Y” problem, not the “I want to execute a complex multi-step DeFi strategy across chains atomically” problem.

What DeFi Protocols Should Actually Do

1. Concentrate liquidity on winning chains. Deploy your core pools on Arbitrum, Base, and Ethereum mainnet. If you have a specific use case on another chain, deploy there too. But don’t spread across 15 chains with thin liquidity on each.

2. Use cross-chain routing for token distribution, not liquidity. Let users bridge to your chain of choice rather than fragmenting your liquidity across every chain that asks for a deployment.

3. Build oracle infrastructure before deploying lending markets. If a chain doesn’t have Chainlink feeds or sufficient DEX depth for TWAP oracles, don’t deploy a lending market there. Unreliable price feeds on thin-liquidity chains are a ticking time bomb.

4. Accept that some DeFi use cases are single-chain by nature. Complex strategies, flash loans, and atomic composability will remain single-chain for years. Design around this constraint rather than pretending cross-chain DeFi is around the corner.

The chain glut didn’t just fragment liquidity — it fragmented the entire DeFi stack. Oracles, liquidation bots, MEV searchers, yield aggregators, and governance systems all need to be replicated on each chain, and most chains can’t support the economics of maintaining all that infrastructure.

Steve’s consolidation thesis is correct: the market will prune. But for DeFi specifically, the damage from the fragmentation period — burned incentive treasuries, failed lending markets, exploited thin-liquidity pools — will leave scars that persist even after the pruning is complete.