Curve Finance launches on Stellar in first native deployment outside EVM ecosystem

Curve's StableSwap invariant has anchored stablecoin trading on Ethereum since 2020, and every prior expansion, from Polygon to Avalanche to various L2s, stayed within EVM-compatible environments where Solidity contracts and veCRV governance ports over directly. Stellar runs on Soroban, a non-EVM smart contract platform the network activated in 2024, built for fast, low-cost payments rather than Ethereum-style composability. This launch is Curve's first attempt to carry its core liquidity design outside that familiar EVM substrate.

Stellar users can now open StableSwap pools, trade compatible assets with low slippage, and supply liquidity directly on mainnet, something previously unavailable without routing through bridges or wrapped EVM assets. Stablecoin issuers and tokenized-asset platforms building on Stellar gain a liquidity venue with Curve's proven pricing curve rather than a thinner, chain-native AMM. FXSwap, the module for foreign-exchange-pegged pairs, stays off-limits until its audit clears, so forex-style markets on Stellar remain unavailable for now.

Curve's relevance has long rested on an EVM-specific flywheel: veCRV bribes, Convex wrappers, and composability with other Solidity protocols. Porting StableSwap to Stellar tests whether that liquidity-and-governance machinery survives transplantation into an account-based chain with a different contract model, or whether Curve ends up running a parallel, thinner version of itself there. The bet implicit in this move is that stablecoin and tokenized-asset liquidity will keep fragmenting across chains rather than consolidating on Ethereum, a premise plenty of DeFi veterans still contest.

The near-term marker is the FXSwap audit: its completion will determine whether Stellar gets Curve's forex-pair functionality or stays limited to plain stablecoin swaps. Beyond that, the number to watch is liquidity uptake in the new Stellar pools, which will show whether users and issuers actually route volume there rather than treating the deployment as symbolic. A second non-EVM rollout, if one follows, would confirm this is a strategy rather than a one-off experiment.