Summary: Skip Go Fast is a solver-financed fast bridge that turns finality delay into a balance-sheet and settlement problem. Users submit an order on the source chain, solvers front assets and optional contract execution on the destination chain within seconds, and only later recover principal plus fees through a cross-chain settlement flow. The useful mechanism lens is simple: speed comes from solvers warehousing reorg and settlement risk for a fee, which makes Go Fast a good foil for Across and proof-heavier interop proposals such as RIP-7755.
What it does:
Lets users submit a fast transfer order from a source chain with assets, destination recipient, optional message payload, and fee terms
Allows permissionless solvers to monitor order events, decide whether the reward/risk profile is attractive, and fulfill the order on the destination chain by fronting capital there
Supports both simple asset delivery and bridge-and-act style execution with an attached message payload
Provides a later settlement path where the solver initiates settlement after fulfillment and receives reimbursement plus fees back on the source side
Ships open-source protocol contracts, an open-source reference solver, and public contract addresses for current supported chains
Exposes timeout/refund behavior so expired unfilled orders can be refunded onchain rather than silently hanging forever
Key claims:
The docs explicitly frame Go Fast as a decentralized bridging protocol that accelerates cross-chain actions by letting solvers front liquidity, which makes it a useful comparison class for any “instant” bridge that is really selling balance-sheet speed rather than cryptographic finality speed
The strongest analytical distinction is between user completion and economic completion: from the user’s perspective the action is done as soon as the solver fills on the destination chain, but from the system’s perspective the hard work continues in the later settlement and rebalancing phases
The official docs say anybody can run a solver and point to a reference solver repo, but the operational docs and solver README make clear that serious participation still requires chain-specific infrastructure, capital inventory, gas management, relaying, settlement monitoring, and rebalancing logic
The contract and solver repositories show that settlement currently relies on a cross-chain messaging system and verification service rather than on a RIP-7755-style destination storage proof, which makes Go Fast a good foil for proof-based interop proposals
The Go Fast docs are explicit that solvers are paid out of the spread between user input and output amounts plus gas-related fees, so the system is structurally a market for temporary capital and execution risk, not just a routing layer
The fee model described in the docs — currently a 10 bps transfer fee plus source and destination gas fees — is useful because it reveals where rent accrues: solver balance sheets, route selection, settlement overhead, and the gateway that decides when speed is worth paying for
The supported-ecosystems docs currently frame Go Fast as an EVM-to-Cosmos speed rail, while other Skip Go API routes rely on CCTP, Axelar, Eureka, or IBC. That separation matters because Skip Go the API/aggregator and Go Fast the protocol should not be collapsed into one abstraction
The failure-cases docs are especially revealing: if no solver responds before expiry, a timeout/refund path triggers. That shows Go Fast is not pretending away failed liquidity provision; it exposes an explicit market-making failure mode rather than only bridge-verification failure modes
The docs also note that enabling Go Fast prioritizes speed over lower fees. That is analytically important because it makes Go Fast a priced latency product for cross-chain users, not merely generic interoperability infrastructure
Whitepaper: The official docs link to a Go Fast whitepaper hosted on Notion, but that document was not retrievable through lightweight fetch tooling in this pass. The strongest accessible primary sources were the official docs plus the open-source contracts and solver repositories; see ../whitepapers/skip-go-fast-primary-sources-2026-05-09.md.