What Is Optimism (OP Mainnet)? The Superchain, OP Stack, and OP Token, Explained

Every rollup promises cheaper Ethereum. Optimism's distinctive bet was never just lower fees — it was that an L2 should be a template other chains copy, not a walled garden. The result is two things at once: OP Mainnet, a top-tier optimistic rollup you can use today, and the OP Stack, the open-source software behind a whole federation of chains called the Superchain. In 2026 that thesis got its hardest test when Base, the Superchain's biggest chain, walked out the door. This is the full picture — how the rollup actually works, what the OP token does, how the Superchain hangs together, and where the real risks sit.

Who built Optimism, and why

Optimism grew out of Plasma Group, a non-profit research team that in 2020 pivoted from Plasma to optimistic rollups and rebranded as Optimism (originally under OP Labs / Optimism PBC). Its founders — including Jinglan Wang, Karl Floersch, Kevin Ho and Ben Jones — came at scaling with an unusually ideological framing: scaling Ethereum was a means, and the end was funding public goods. That belief is baked into the org chart. Optimism is governed by the Optimism Collective, and a fifth of the token supply is set aside to pay back the people who build valuable open-source infrastructure.

OP Mainnet went live in late 2021 (initially as "Optimistic Ethereum"), making it one of the first EVM-equivalent rollups to reach production. "EVM-equivalent" is the specific claim worth understanding: rather than a custom VM that behaves like Ethereum, Optimism aimed to run the actual Ethereum execution environment with minimal changes, so contracts and tooling port over with almost no friction. That design choice is why so much of the ecosystem was able to deploy quickly, and it's the seed of the OP Stack.

The architecture: rollup type, proof, data, sequencer

Four things define any rollup. Here's where OP Mainnet stands on each in 2026.

Rollup type — optimistic, not ZK. OP Mainnet batches transactions, executes them off-chain, and posts the compressed data plus resulting state roots to Ethereum. It assumes each batch is valid — hence "optimistic" — and only checks the math if someone disputes it. This is the opposite of a ZK rollup, which attaches a cryptographic validity proof to every batch up front. Optimism's approach is simpler and cheaper to run, at the cost of a challenge window before withdrawals to L1 finalize.

Proof system — Cannon, now permissionless. The historically weak spot of optimistic rollups was that fault proofs weren't actually live, so users trusted a multisig. That changed in June 2024, when Optimism shipped permissionless fault proofs using Cannon, its fault-proof virtual machine. Any participant — not just OP Labs — can now challenge an invalid state root and block a fraudulent withdrawal. This pushed OP Mainnet to L2BEAT Stage 1, the maturity tier where fraud proofs are real and permissionless. The roadmap goes further: the fault-proof system is modular by design, and Optimism is developing additional proof systems (Asterisc and Kona, and ZK proving) toward a multi-proof setup that reduces reliance on any single implementation — a step toward Stage 2.

Data availability — Ethereum, via blobs. OP Mainnet posts its transaction data to Ethereum itself, and since the Dencun upgrade it uses EIP-4844 blobs, which slashed data-posting costs and, with them, user fees. Posting to L1 is what gives a rollup its security: anyone can reconstruct and verify the chain from Ethereum data. (The OP Stack also supports "alt-DA" modes for cheaper chains, but OP Mainnet uses Ethereum.)

Sequencer — still centralized. This is the honest caveat. The sequencer — the node that orders transactions and produces L2 blocks — is operated centrally by OP Labs. It can't steal funds (fault proofs and L1 settlement prevent that), but it can reorder or censor transactions in the short term, and if it goes down the chain stalls. Decentralizing sequencing — and eventually sharing it across the Superchain — is a headline roadmap item, tied to the interop work, but it is not done. A Security Council can also intervene under defined emergency conditions, which is a trust assumption you should price in.

Fees, speed, and performance

For everyday use, OP Mainnet feels like a fast, cheap Ethereum. Transactions confirm on the L2 in about two seconds (soft confirmation from the sequencer), and typical fees run from a fraction of a cent to a few cents, depending on Ethereum's blob market and how busy the chain is. Since blobs arrived, the dominant cost of an L2 transaction — posting data to L1 — dropped sharply, so fees are usually trivial for swaps and transfers.

Two nuances matter. First, finality is layered: your transaction is soft-confirmed instantly by the sequencer, but true Ethereum-backed finality follows once the batch settles on L1. For most on-chain activity the soft confirmation is what you interact with. Second, withdrawals to Ethereum go through the challenge window — historically about seven days — because that's the time during which a fault proof can dispute the exit. Deposits into the L2 are fast; getting back out the native way is deliberately slow. Third-party "fast withdrawal" bridges exist that front you liquidity immediately for a fee, sidestepping the wait.

The OP token and how incentives work

Yes, Optimism has a token, and it's real: OP, an ERC-20 launched in 2022 with an initial supply of 4.29 billion. Two things people get wrong about it:

• OP is not the gas token. You pay gas on OP Mainnet in ETH, like most Ethereum rollups. OP is a governance and incentive asset, not a fee currency.
• A large slice is reserved for public goods. Around 20% of supply (~850M OP) is earmarked for public-goods funding — the mechanism Optimism is arguably best known for.

OP's main job is governing the Optimism Collective, and it's used for ecosystem incentives — liquidity programs, grants to chains and apps, and reward campaigns — that bootstrap activity across the Superchain. Its market value tracks the fortunes of the whole ecosystem, which is exactly why it dipped when Base announced its exit in early 2026.

Retro Funding and the Optimism Collective

The Collective runs on a bicameral design. The Token House (OP holders) submits and votes on governance proposals; the Citizens' House (a set of designated citizens) allocates public-goods rewards. The flagship program is Retroactive Public Goods Funding, now run as Retro Funding.

The idea is a genuinely original contribution to crypto governance: instead of guessing which projects deserve grants before they build (the usual VC-style bet), you reward impact after it's demonstrably happened. It's easier to agree that something was useful than to predict that it will be. Optimism has distributed tens of millions of OP across multiple rounds this way, funding infrastructure, tooling, governance research and developer education — and in 2025–2026 it moved from discrete annual rounds toward an ongoing impact-evaluation model with more regular payouts. Whether retroactive funding scales cleanly is still debated, but it has been copied across the industry, which tells you it struck a nerve.

The Superchain and the OP Stack

Here's what separates Optimism from a plain rollup. The OP Stack is the open-source, MIT-licensed software that runs OP Mainnet — and anyone can use it to launch their own L2. The chains that adopt it and agree to a shared set of standards, security assumptions and (eventually) a communication layer form the Superchain.

By 2026 the Superchain spanned roughly a dozen production chains, historically including World Chain (from the Worldcoin/World project), Unichain (Uniswap's L2), Zora, Mode, and — until early 2026 — Base. Governance of the shared standards runs through the Law of Chains, a framework meant to define the rights and expectations of chains, users and operators in the Superchain so that no single actor can unilaterally change the rules. The endgame is interoperability: native, near-instant movement of assets and messages between Superchain chains, plus shared/decentralized sequencing — turning many chains into something closer to one horizontally scaled network. Interop has been progressing on testnet, with fuller cross-chain composability targeted for later-2026 upgrades.

The caveat, and 2026's defining event: in February 2026, Coinbase announced Base would migrate off the shared OP Stack to its own Base-operated codebase, citing a desire to control upgrades without coordinating across multiple teams. Base was the Superchain's largest chain by TVL (around $3.85B at the time), so its departure was a genuine blow to the "one Superchain" narrative and OP's price. It doesn't unwind OP Mainnet or the other chains, but it's a live reminder that the Superchain's strength is only as durable as its members' willingness to stay.

The ecosystem: apps and liquidity

OP Mainnet itself is a mature DeFi and app chain. Its TVL has swung between several hundred million and around $1B and took a real hit after Base's 2026 exit — still a top-tier L2, though behind Arbitrum individually. You'll find the usual blue-chip crowd deployed: Uniswap, Aave, Velodrome (the native ve(3,3) DEX that anchored OP's early liquidity), Synthetix (which originated much of the perps/derivatives activity here), plus lending, bridges and stablecoins. Because OP Mainnet is EVM-equivalent, almost anything on Ethereum runs here with minimal changes, so the app selection is broad and familiar rather than exotic.

The more interesting number is the aggregate Superchain: sum OP Mainnet, World Chain, Unichain, Zora, Mode and the rest, and the combined footprint has at times been competitive with Arbitrum. That aggregation is the whole pitch — and also why losing Base stung.

How to get on: bridging and wallets

Getting onto OP Mainnet is routine:

1. Add the network. Any EVM wallet (MetaMask, Rabby, a smart-account wallet) works. Add OP Mainnet — most wallets have it built in — or connect and let a dApp prompt the switch.
2. Bridge assets. Use the official bridge at app.optimism.io/bridge or an aggregator like Superbridge to move ETH and tokens from Ethereum. Deposits arrive in minutes. You keep gas in ETH, not OP.
3. Withdraw carefully. Native withdrawals to L1 pass through the challenge window (about a week). For speed, a third-party fast bridge will front you liquidity for a fee — a convenience, and a small extra trust assumption.
4. Move within the Superchain. Between OP Stack chains, purpose-built bridges (and, increasingly, interop) make hops cheaper and faster than round-tripping through Ethereum.

Only bridge what you intend to use, and reach bridges through your own bookmarks — fake "bridge" front-ends are a perennial phishing vector.

Risks and what to watch

A clear-eyed list:

• Centralized sequencer. OP Labs still runs sequencing. It can't steal funds, but short-term censorship or downtime is possible until decentralized sequencing ships. This is the biggest gap between "Stage 1" and "trustless."
• Security Council backstop. A council can intervene under defined conditions — a pragmatic safety measure, but also a trust assumption. Full Stage 2 (minimal governance override) is a roadmap goal, not today's reality.
• Superchain cohesion risk. Base's 2026 exit is the case study: chains can leave, and the network's value depends on them choosing to stay and on interop actually shipping. Watch whether interop lands and whether other large chains recommit or defect.
• Withdrawal delay. The ~7-day native exit is inherent to the optimistic design. Fast bridges reduce the friction but add counterparty risk.
• Token/governance risk. OP is volatile, and retroactive public-goods funding — however admired — is still an experiment in allocating large sums without capture. Governance disputes over who deserves funding have already flared.

Bottom line

Optimism is two things worth keeping straight. OP Mainnet is a mature, EVM-equivalent optimistic rollup with live permissionless fault proofs, cheap blob-based fees and a deep app ecosystem — a solid, top-tier place to transact, with the honest asterisks of a centralized sequencer and a slow native exit. The OP Stack and Superchain are the bigger, riskier bet: an attempt to make the winning L2 a shared standard rather than a single chain, funded by a genuinely novel public-goods engine. Base's 2026 departure showed that bet is contestable. But permissionless fault proofs, a real governance token, and a working retroactive-funding model are more than most L2s can claim — and they're why Optimism still sets a lot of the agenda even when its biggest tenant walks away.

For related reading: best Ethereum L2s, what is DeFi, and account abstraction explained.

Related guides

• Arbitrum
• Starknet
• Status Network
• Robinhood Chain

Not financial advice. OP is volatile, rollup mechanics and Superchain plans can change — always verify details on official Optimism channels.