Layer 2 coins on a live bubble chart

Every coin on this page belongs to a network that runs on top of another blockchain. Arbitrum, Optimism, Starknet, ZKsync, Mantle and the rest of the layer 2 coins process transactions on their own, then settle back to Ethereum (or, for a few, to Bitcoin) and borrow its security. The bubbles above show how their tokens are moving right now, sized and colored so you don't have to read a price table line by line.

It's a filtered slice of the main chart. The home page runs the same altcoin bubble chart across the whole market, 1,000+ coins at once. Here, only the scaling networks are left, which matters more than it sounds: layer 2 tokens are small next to the chains they're built on, and on the full chart they'd disappear behind Ethereum.

What the layer 2 crypto bubbles are telling you

Start with color. Green means a token is up over the timeframe you picked, red means down, and brightness tells you by how much. Layer 2 tokens move as a pack more than almost any other category on this site, because most of them are really one bet: that Ethereum keeps getting used. When the whole page turns the same color, it's usually ETH's mood spilling over and not news about any single network.

The split days are the interesting ones. If ARB and OP are flat and Starknet or ZKsync is glowing green on its own, something specific happened (an upgrade, a big app launching there, an exchange listing, a new airdrop season). The reverse is worth noticing too. One red bubble in an otherwise green pack is often a token unlock landing, which gets its own section further down.

Size depends on the mode. Performance mode, which I leave on most of the time, sizes bubbles by how far they moved, so a small token that jumped 20% can look bigger than Arbitrum for a day. Market cap mode is more honest about the pecking order, and a bit harsh: a handful of tokens take most of the screen, and a long tail of rollups nobody talks about anymore sits in tiny circles at the edges.

For timeframes, I mostly use the 7d view here. The 24h view on layer 2s is jumpy, since these tokens react to every Ethereum headline and to Bitcoin's mood on top of that. The 1y view is the brutal one. It shows how far a lot of layer 2 tokens have fallen from their launch prices, which is useful context before you believe any single green week.

What is a layer 2, in plain terms

The textbook version is that a layer 2 executes transactions away from the main chain and posts the results back to it. That's correct. It also skips why anyone bothered.

Ethereum can only fit so much into each block. In 2021, when DeFi and NFTs were at full speed, a simple token swap on Ethereum could cost $50 or more in fees. Layer 2s were the answer Ethereum settled on: move the everyday activity to separate networks, bundle thousands of transactions together, and post only a compressed summary back. Ethereum stays the final record, and the layer 2 inherits its security for as long as its proofs and bridge hold up.

The turning point came on March 13, 2024, with Ethereum's Dencun upgrade. It added a cheaper kind of data space, called blobs, built specifically for rollups, and layer 2 fees fell by 90% or more almost overnight. A simple transfer on the big networks now usually costs a cent or less. That's also when the relationship between layer 2s and ETH got complicated, which the layer 1 vs layer 2 section on the layer 1 page covers from Ethereum's side.

Here's what most explainers leave out. On most Ethereum layer 2 networks, you don't need the network's own token to use it. Gas on Arbitrum, Optimism and Base is paid in ETH, and ARB and OP are governance tokens. Starknet is an exception, since STRK can pay for fees, and Mantle uses MNT for gas. That one design choice explains a lot of what you'll see on this chart.

Optimistic rollups vs ZK rollups

Almost every layer 2 on this page is a rollup, and rollups come in two kinds. The optimistic rollup vs ZK rollup difference comes from one question: how does Ethereum know the layer 2 isn't cheating?

An optimistic rollup assumes every batch is honest and gives anyone about seven days to prove otherwise with a fraud proof. That week is why withdrawing from Arbitrum or Optimism through the official bridge takes so long. Almost nobody actually waits, though. Third-party bridges pay you out on Ethereum in minutes and sit through the challenge window themselves, for a small fee.

A ZK rollup posts a validity proof with every batch, a piece of cryptography showing the transactions were computed correctly. Ethereum checks the proof and that's the end of it, with no waiting period. The catch has been complexity. ZK proofs are hard to build and expensive to generate, which is why the optimistic rollups reached the market first and still carry most of the activity.

For reading the chart, the rollup type almost never moves a price by itself. Adoption, unlocks and ETH's direction do that. Where it matters is if you actually hold funds on one: how long a withdrawal takes, and what's really backing the bridge.

A layer 2 crypto list, with the short version of each

Think of this as a short layer 2 crypto list: the networks that usually end up as the biggest bubbles here, plus two that never show up at all. Years are mainnet launches, and the last column is my own one-line summary, so feel free to disagree with it.

Base is the row that trips people up. It's one of the busiest layer 2s, but it launched without a token of its own, so for most of its life its activity hasn't shown up on this chart at all. That's the single biggest gap between what's happening on layer 2s and what the bubbles can show. The Stacks row starts arguments too, since whether anything built on Bitcoin counts as a real layer 2 depends on who you ask. And POL sometimes appears here as well, for reasons the FAQ gets into.

Why layer 2 tokens and layer 2 usage drift apart

If you only take one thing from this page, make it this one: a layer 2 can double its users while its token goes nowhere.

The first reason is the fee design from earlier. When gas is paid in ETH, more activity doesn't create demand for ARB or OP. The token gets you votes in a DAO, and markets haven't paid much for votes. Optimism has tried to change that with the Superchain, where chains built on the OP Stack, Base included, share part of their sequencer revenue with the Optimism Collective. Whether that ever reaches the OP price in a way the market cares about is still an open question.

The second reason is supply, and it's the one that shows up on the chart as a sudden red bubble. Most layer 2 tokens launched with a small circulating float and large allocations for teams and early investors, released on a schedule. Arbitrum's first big cliff unlock, on March 16, 2024, released about 1.1 billion ARB, a huge jump in circulating supply in a single day. Starknet chose a steadier monthly schedule of about 64 million STRK instead. Either way, new supply keeps arriving whether or not the network is growing.

So before reading much into a bubble, check the token's unlock calendar and compare its market cap with its fully diluted valuation. When you compare tokens side by side, look at circulating supply against total supply first. A token with only a quarter of its supply circulating has a lot of selling ahead of it, however good the tech is.

The layer 2s that faded (and one that got switched off)

Layer 2 history is short, but it already has a graveyard.

The first serious attempt at scaling Ethereum wasn't rollups at all. It was Plasma, around 2017 and 2018, with OMG Network as its best-known project. Plasma chains kept their data off Ethereum and relied on users watching for fraud and exiting in time, which proved clumsy for anything beyond simple payments. Rollups replaced the idea, and OMG faded along with it. Nothing exploded. The design just lost.

Polygon zkEVM is the most recent case, and it's worth knowing because it wasn't a failed startup. It came from one of the biggest teams in crypto, launched in 2023 as a flagship ZK rollup, and never attracted much use. Polygon announced in June 2025 that it would sunset the chain, and the sequencer was shut down in July 2026. People with funds still there can claim them on Ethereum until the end of 2027. If you ever wondered whether a layer 2 can simply be switched off, that's your answer.

Then there's the airdrop cycle. Several layer 2s launched with points programs, headline deposit numbers and a token at the end, and a lot of that activity left as soon as the airdrop landed. Blast is the usual example: it pulled in billions of dollars in deposits in 2024 by paying yield on bridged funds, then saw its activity and token price fall hard after launch. On the bubble chart this looks like a token that starts big and gets a little smaller every month, which the 1y view shows better than any other setting.

The common thread is that a layer 2 fades when people stop having a reason to be there, and a token doesn't fix that. Cheap fees, a bridge people trust and one or two apps worth using keep a network alive. Points programs mostly rent users for a while.

Layer 2 vs layer 1, and where the line gets blurry

The clean version: a layer 1 runs its own consensus and secures itself, while a layer 2 executes transactions separately and leans on a layer 1 for final settlement and security. Ethereum, Solana and Bitcoin live with the other layer 1 coins. Arbitrum, Optimism and the rest are here.

The blurry cases are where the arguments happen. Sidechains come first. Polygon PoS has its own validators and only checkpoints to Ethereum, so it doesn't inherit Ethereum's security the way a rollup does. Some data sites still file POL under layer 2, which is why it can turn up on this chart. I'd call it a sidechain and move on.

Validiums and so-called alt-DA chains are the next grey zone. They prove their transactions to Ethereum but keep the underlying data somewhere cheaper, so if that data ever went missing you couldn't necessarily rebuild everything from Ethereum alone. A lot of gaming and high-volume chains work this way. They're cheaper, and a little less of a layer 2 than the name suggests.

Bitcoin layer 2 networks are their own argument. Bitcoin can't verify rollup proofs the way Ethereum can, so most projects using the label rely on federations, sidechains or new trust assumptions of some kind. Lightning is the cleanest case: payment channels that settle on Bitcoin, with no token at all. Stacks anchors its blocks to Bitcoin, went through its Nakamoto upgrade in late 2024 and then launched sBTC to bring BTC into its apps. Rootstock is a sidechain secured by merged mining. Plenty of newer projects call themselves Bitcoin layer 2s mostly because the label sells.

And then there's layer 3: chains that settle to a layer 2 instead of directly to Ethereum, like some of Arbitrum's Orbit chains. The numbering will keep going as long as it helps with marketing.

What this page won't tell you

It's a price and momentum view of layer 2 tokens. That's the whole job.

It can't show activity. Transactions, active addresses and fees per network are better on L2BEAT's activity pages or on growthepie, and both are free. Given how loosely token prices and usage line up in this category, it's worth keeping one of them open next to the bubbles.

It can't show risk either. Two rollups can look identical as circles while one has working fraud proofs and a security council with real limits, and the other can be upgraded overnight by a small team. L2BEAT's stage ratings (Stage 0, 1 and 2) are the standard way to check, and I'd look before bridging real money anywhere.

And it can't show networks without a token, which leaves out some of the busiest layer 2s entirely. The list itself follows CoinGecko's layer 2 category tagging, so edge cases like POL inherit whatever call CoinGecko made. We don't hand-curate it, and it seems better to say so than to imply otherwise.

A Monday-morning way to use it

Say it's Monday, the 7d view is on, and ZKsync and Starknet are both bright green while ARB and OP sit pale red. Before buying into a "ZK rollups are back" story, click each green bubble and check whether volume rose with the price, because a move on thin volume fades more often than it holds. Then look at the unlock calendars for ARB and OP. A pale red week for the two biggest tokens is sometimes nothing more than a scheduled unlock.

Last, load ETH and the two green tokens into the Compare page, putting ETH first, since correlation is measured against whichever coin you add first. If both come back close to 1.0, the green week was mostly Ethereum lifting everything around it, and the ZK story you were about to believe probably isn't the reason.