The story starts before the Synthetix name existed. Havven launched as a decentralized stablecoin system, with HAV acting as the collateral token and nUSD as its stable asset. In November 2018 the project announced its transformation into Synthetix: HAV became SNX, nomins became Synths and nUSD became sUSD.
The Synth era expanded the original stablecoin idea into synthetic exposure to currencies, commodities, indices and cryptoassets. SNX collateral and a shared debt pool sat underneath that system, creating a distinctive model where stakers collectively carried exposure to the protocol’s synthetic liabilities.
That debt-pool design also made staking unusually demanding. SNX stakers historically had to manage collateral ratios and debt while the value of the system’s liabilities moved. Later redesigns increasingly separated SNX participation from the old requirement to actively manage synthetic debt.
Perps V2 marked another major shift. Synthetix combined skew-sensitive pricing, funding-rate velocity and signed offchain oracle updates to improve perpetual-futures execution while keeping market logic tied to smart contracts. The product was becoming less about a catalogue of Synths and more about derivatives infrastructure.
Perps V3 pushed that architecture toward account-based cross margin and configurable collateral. Its design allowed one account’s margin to support positions across markets and introduced a broader framework for collateral, permissions and integration on optimistic EVM rollups.
The next pivot was more dramatic. During the 2025 rebuild Synthetix chose to deprecate its L2 AMM strategy and concentrate on one Ethereum Mainnet perpetuals venue. The project explicitly framed multi-chain fragmentation, bridge friction and the limitations of AMM-style perps as reasons to change direction.
The Mainnet architecture is deliberately hybrid. A high-performance central limit order book performs offchain matching because Ethereum cannot currently provide the latency and throughput expected from a professional matching engine. User collateral and trade settlement remain on Ethereum Mainnet, separating execution speed from the settlement layer.
That distinction matters. Calling the current product simply an onchain order-book exchange would be inaccurate. Matching is offchain; settlement is onchain. Synthetix has described moving more functionality onchain as Ethereum scales as a longer-term decentralization direction.
By May 2026 Synthetix reported that Mainnet Perps was live and operating through a private-alpha phase. Its trading stack included market and limit execution, linked and standalone take-profit/stop-loss tooling, public integration endpoints and SDK work, while additional order types such as TWAP and scaled orders were still described as upcoming at that checkpoint.
Multi-collateral changes the capital model around the exchange. On 3 June 2026 Synthetix announced ETH as the first non-USDT collateral live on the Mainnet venue, allowing traders to maintain ETH exposure while using it to support USDT-settled perpetual positions.
Liquidity and liquidation are another layer. The Synthetix Liquidity Provider was already acting as market maker and liquidator in the private-alpha system by the May 2026 roadmap update, but public access was still described as a future rollout. A delivery record should keep that distinction visible rather than converting a target into a completed milestone.
SNX staking has also been rewritten. SIP-420, created in 2025 and marked Implemented, introduced protocol-owned debt and delegated staking to reduce the complexity of individual debt management and increase capital efficiency.
Then the direction changed again. SIP-423, created on 12 June 2026 and marked Implemented, calls for retiring legacy sUSD and restructuring the 420 Pool. Crucially, the proposal itself labels the receipt-contract build and the new SNX staking-reform contract as deferred. Governance status and completed software delivery are therefore not identical.
This is why Synthetix is a useful structure case. A protocol can have real technical history, ship a new exchange and still change its economic architecture repeatedly. For SNX, the durable question is whether real trading activity, liquidity, token utility and supply behavior eventually settle into a structure that persists after each narrative reset.