Coinbase's Layer 2 network Base announced on February 18, 2026, that it will abandon the Optimism OP Stack and migrate to a proprietary "unified stack" under a consolidated base/base GitHub repository. The move strips Optimism of its single largest revenue contributor — a chain that generated 71%...
"We're evolving our foundational software by moving to a unified, Base-operated stack to accelerate innovation, scaling and security." — Wilson Cusack, Base Engineering
Coinbase's Layer 2 network Base announced on February 18, 2026, that it will abandon the Optimism OP Stack and migrate to a proprietary "unified stack" under a consolidated base/base GitHub repository. The move strips Optimism of its single largest revenue contributor — a chain that generated 71% of all Superchain sequencer fees in 2025 and accounted for 90% of revenue by January 2026.
The OP token crashed 28% to an all-time low of $0.12 within 48 hours. At $0.14, OP trades 97% below its March 2024 peak of $4.85. The Superchain's remaining roster — Sony's Soneium, Uniswap's Unichain, Kraken's Ink, Worldcoin's World Chain, and OP Mainnet — collectively generated $4.2 million in H2 2025 sequencer revenue, compared to Base's $35.7 million over the same period.
The departure raises a structural question for Ethereum's Layer 2 ecosystem: whether shared infrastructure alliances can survive once a dominant participant outgrows the collective.
Wilson Cusack published the technical rationale on the Base Engineering Blog on February 18, 2026. The core change: Base will consolidate its sequencer, execution client, and consensus components into a single repository (base/base), replacing the distributed dependency model where it relied on code maintained by Optimism, Flashbots, and Paradigm.
Three hard forks are planned under the new architecture, culminating in Base V1, which will align with Ethereum's upcoming Glamsterdam upgrade in mid-2026. Base intends to double its upgrade cadence from three to six protocol upgrades per year.
The governance layer is also decoupling. Base is restructuring its Security Council, replacing seats previously tied to Optimism with independent signers. Node operators will be required to migrate to the new Base client over the next few months, though existing RPCs — including those in the Optimism namespace — will remain supported during the transition.
Base stated it will "continue to work with Optimism" and maintain compatibility with OP Stack standards. The market interpreted this as diplomatic language for an operational separation.
Under the original 2023 agreement, Base committed to sharing either 15% of gross profit or 2.5% of total sequencer revenue — whichever was greater — with the Optimism Collective. In return, Base received up to 118 million OP tokens ($175 million at issuance) vesting over six years, capped at 9% of total votable supply.
The numbers tell a lopsided story:
| Metric | Value | |--------|-------| | Base's 2025 chain revenue | $74 million | | Base's share of Superchain sequencer fees (2025) | 71% | | Base's share of Superchain revenue (Jan 2026) | 90% | | Total ETH contributed to Optimism Collective | 8,387 ETH (~$16.4 million) | | Base's contribution as % of Optimism's lifetime revenue | 41% | | Base's effective rent payment | 2.5% of revenue |
Base generated $74 million in chain revenue during 2025 but paid roughly 2.5% — approximately $1.85 million — back to the Optimism Collective. The 15% gross profit trigger was designed for a scenario where Base operated at thinner margins; in practice, the 2.5% floor became the operative rate.
For Optimism, the loss is existential in revenue terms. In H2 2025, Base generated $35.7 million in sequencer revenue. The next largest contributor was Unichain at $2.7 million, followed by OP Mainnet at $1.5 million. The remaining chains — World Chain ($1.2 million implied by gas share), Ink, and Soneium — contributed proportionally less.
Total H2 2025 Superchain sequencer revenue allocated to the Optimism Collective was $7.2 million, already a 13% decline from $8.3 million in H1. Without Base, that figure will contract further.
Base's operational metrics explain both its leverage and its motivation to depart:
At this scale, the OP Stack's shared dependency model became a constraint rather than an enabler. Base's engineering team was limited to three protocol upgrades per year — a cadence dictated by the OP Stack's release cycle. Under the unified stack, Base targets six annual hard forks.
The sequencer fee controversy also played a role. Blockchain analysts had previously flagged that Base was routing sequencer fees off-chain to Coinbase custody. The Base team defended the practice as necessary "for security and audit purposes," but the optics reinforced the perception that Base operated as a Coinbase proprietary chain wearing Superchain clothing.
The Superchain reached 34 OP Chains by the end of H2 2025, with total transactions hitting an all-time high of 3.6 billion — up 44% from H1. But the distribution of value was never uniform.
Without Base, the Superchain's remaining revenue profile:
| Chain | H2 2025 Sequencer Revenue | Gas Share | |-------|--------------------------|-----------| | Unichain | $2.7 million | ~3.4% | | OP Mainnet | $1.5 million | ~11.4% | | World Chain | — | ~11.5% | | Ink | — | ~4.7% | | Soneium | — | ~4.2% |
The remaining chains have credible backers — Sony (Soneium), Uniswap (Unichain), Kraken (Ink), Worldcoin (World Chain) — but none approaches Base's revenue generation. The Superchain's value proposition now rests on shared interoperability standards and governance rather than on a revenue-generating anchor tenant.
According to Messari's H2 2025 State of the Superchain report, revenue allocated to the Optimism Collective was already declining before Base's announcement. The post-departure trajectory points to a Superchain that functions more as a technical standard than as an economic alliance.
The Optimism Collective moved preemptively. On January 30, 2026, governance approved a proposal to allocate 50% of net Superchain sequencer revenue to a 12-month pilot of monthly OP token buybacks. Purchased tokens will be held in the Collective treasury for future governance-directed use.
The buyback mechanism serves two purposes: provide a price floor for OP by reducing circulating supply, and signal that the Collective can generate value independent of any single chain. Whether the remaining revenue base — potentially under $5 million annually without Base — is sufficient to make buybacks material is an open question.
OP whales have been buying the dip, according to on-chain data cited by CCN. The thesis appears to be that OP at $0.14 — 97% off its peak — prices in a worst-case scenario, and that the remaining Superchain ecosystem plus new entrants could stabilize revenue.
Base's departure from the OP Stack mirrors a broader trend: the largest L2s are vertically integrating rather than relying on shared infrastructure.
Arbitrum built its own Nitro stack from the outset. zkSync developed its own ZK prover. StarkNet maintains a proprietary Cairo VM. Base's move brings Coinbase's L2 into alignment with this pattern — the dominant L2 players all control their own execution environments.
The implications extend beyond Optimism:
Shared stack economics are fragile. The Superchain model assumed aligned incentives would keep participants contributing. Base's 2.5% rent — on $74 million in revenue — was not enough to prevent departure once the engineering constraints became binding.
L2 consolidation accelerates. Base's TVL dominance (46% of L2 market) combined with its corporate backing creates a flywheel effect. Coinbase's 110+ million verified users serve as a distribution moat that pure-play L2s cannot replicate.
Token value for infrastructure DAOs is under pressure. OP's 97% drawdown from peak reflects a market reassessment of what governance tokens for shared infrastructure are worth when the infrastructure's largest user can walk away.
This development arrives as the broader L2 sector faces headwinds. According to The Block's 2026 Layer 2 Outlook, the proliferation of chains has diluted liquidity across the ecosystem. The Ethereum Foundation's response — the Glamsterdam hard fork with gas limit increases to 100–200 million — aims to make L1 competitive again, adding another source of existential pressure on L2 business models.
Base's departure from the OP Stack is not a partnership dispute. It is the predictable outcome of an economic structure where the largest participant captured 71% of revenue, paid 2.5% rent, and was constrained by a shared upgrade schedule designed for the median contributor.
For Optimism, the path forward requires either attracting a new anchor tenant of comparable scale — an unlikely prospect given Base's unmatched Coinbase distribution channel — or repositioning the Superchain as a lightweight technical standard rather than an economic alliance. The OP buyback program buys time but does not solve the fundamental revenue gap.
For the broader L2 ecosystem, Base's move clarifies an emerging rule: shared infrastructure works for bootstrapping, but dominant chains will eventually internalize their stacks. The question is whether Ethereum's own L1 scaling — via Glamsterdam's gas limit increases — will render the entire L2 revenue model obsolete before any of them can build a durable moat.