Bridge speed is the gap between clicking "send" on chain A and your funds being spendable on chain B. In 2026 that gap runs from the 1.2-second average Across publishes for an intent fill to the 15 to 19 minutes Circle documents for a CCTP Standard Transfer out of Ethereum, and the right answer depends entirely on which route you pick. This guide collects the speed figures each protocol publishes, explains why finality models drive them, and helps you choose the fastest route per destination.
​

Fastest means time from source-chain inclusion to destination-chain spendability, not just the source-chain confirmation. A bridge that confirms in 2 seconds but waits 15 minutes for finality before minting on the destination is not fast. The end-to-end clock is what users feel.
​

Three things gate that clock: source-chain finality requirements, the bridge's verification model (optimistic, intent-based, or canonical), and destination-chain block time. Intent bridges like Across short-circuit finality by fronting liquidity from a relayer; canonical bridges like CCTP wait for source finality before minting native USDC.
​

The figures below are the ones the protocols themselves publish. Where a protocol does not publish an end-to-end time, this guide does not print one: Stargate, LayerZero and Hyperlane document their verification models rather than per-route delivery times, so no number is listed for them.

Two patterns jump out. First, intent bridges that front liquidity are an order of magnitude faster than an attestation that waits for hard finality. Second, the source chain sets the floor: Circle's own table shows a Standard Transfer taking 15 to 19 minutes out of Ethereum and OP Stack rollups but about 8 seconds out of Polygon PoS and Avalanche, because those chains finalize on their own schedule rather than Ethereum's.
​

CCTP burns USDC on the source chain and mints native USDC on the destination. A Standard Transfer waits for hard finality on the source chain, which Circle CCTP docs puts at 15 to 19 minutes out of Ethereum and about 8 seconds out of Polygon PoS. A Fast Transfer attests after inclusion instead, in roughly 8 to 20 seconds, at the cost of a global allowance that caps reorg exposure. Across skips the wait entirely by having a relayer pre-fund the destination and reconcile later. You get speed; the relayer takes finality risk and prices it into the fee.
​

This is the speed-versus-trust trade. Canonical bridges return native assets and inherit only Ethereum-level trust. Intent bridges return native assets too (Across fills with the canonical asset) but layer in a relayer assumption that gets unwound after L2-to-L1 finality. For most users moving under $50k, the trade is worth it.
​

Every bridge benchmark above is downstream of source-chain finality. Here is what each chain actually guarantees:

L2BEAT tracks these finality assumptions and the trust models behind every rollup, and is the canonical reference for which "soft" finality numbers you can rely on.
​

Relayers fill destination orders and reconcile with the canonical bridge later. Fastest for ETH and major stables on EVM L2s. across.to publishes an average of 1.2 seconds from confirmation to funds in the wallet, with most transfers landing in under two seconds.
​

Unified liquidity pools with delta algorithm for rebalancing. Faster than canonical, slower than intent. Best for routes where you want a single pool of liquidity and finality-aware messaging. Source: stargateprotocol.gitbook.io.
​

General-purpose messaging layer. Token bridges built on LayerZero (USDT0, OFTs) inherit DVN-based verification times, typically 60-120 seconds end-to-end. Source: docs.layerzero.network.
​

Modular interchain security model with warp routes for token transfers. Faster on routes where validator sets are well-staffed. Eco partners with Hyperlane for cross-chain message delivery. Source: docs.hyperlane.xyz.
​

Burns USDC on source, attests via Circle, mints native USDC on destination. A Standard Transfer waits for source-chain hard finality; a Fast Transfer attests in roughly 8 to 20 seconds against a global allowance. Either way the destination receives native USDC with no liquidity pool in the path. Source: Circle CCTP docs.
​

Relayer inventory exhaustion. Intent bridges are only fast while the relayer has destination liquidity. Large transfers can fall back to canonical-bridge speed without warning. Always check the bridge's quoted fill time before sending.
​

Source-chain congestion. Even a 10-second intent bridge becomes a 60-second bridge if the source chain has full blocks. Include source-chain inclusion time when budgeting.
​

Destination chain block time. A bridge that submits quickly still has to wait for the next destination block. Fast destination chains help: Solana activated a mainnet feature gate cutting slot time to 300ms in August 2026, per the Solana changelog, August 27 2026. Ethereum's 12-second blocks hurt.
​

Eco Routes is an intent-based protocol that quotes execution across multiple underlying bridges (Hyperlane for messaging, CCTP for canonical USDC, plus integrations with major liquidity layers) and picks the fastest viable path for your transfer size. You submit one intent; solvers compete to fill it on the destination in seconds. Speed becomes the default, not a manual optimization.
​

Every time in this guide is a figure the protocol itself publishes, read from the linked page on September 16, 2026. No per-route median is claimed for protocols that do not publish one, and no figure here comes from private testing. Quoted times describe typical conditions; relayer inventory, source-chain congestion and transfer size all move them, so check the quote shown at send time.
​

Sources: across.to, Circle CCTP docs, Solana changelog, August 27 2026, and l2beat.com.
​