What is the cheapest crypto bridge in 2026?

It depends on the route, which is why aggregators exist. Across is typically cheapest for USDC between Ethereum L2s, settling in the 0.04% to 0.10% range with sub-minute latency. Stargate is competitive at 0.01% to 0.06% for native asset delivery on supported pools. Across and deBridge tend to produce the tightest cost on ETH and WBTC. Rather than memorising this, quote an aggregator like Jumper on your exact pair and size, because the cheapest route changes with liquidity conditions.

Are crypto bridges safe?

Bridges hold the record for the largest exploits in crypto history, with several individual incidents exceeding $300 million. The modern intent-based designs used by Across, Relay and deBridge are structurally safer than the old lock-and-mint model because they hold far less pooled capital at any moment: a relayer fronts your funds and is reimbursed afterwards. Safer is not safe. Bridge what you need, when you need it, rather than treating a bridge as storage.

Should I use a bridge or an aggregator?

An aggregator, unless you bridge the same route constantly and already know which venue wins it. Jumper, built on LI.FI, quotes Across, Relay, Stargate, Mayan and others live and routes to the best combination of cost, speed and slippage. The small routing fee is almost always less than the difference between the best and worst route on any given day.

How long does bridging take?

Intent-based bridges settle in seconds to about a minute on major EVM routes, because a relayer fronts the funds on the destination chain immediately and gets reimbursed later. Pooled-liquidity bridges like Stargate are near-instant on supported pools. Routes through Cosmos via Axelar take minutes. Anything advertising a multi-day challenge period is a canonical rollup bridge, which is the slowest and also the most trust-minimised option.

What is the safest way to bridge a large amount?

Split it. Send a small test transaction first, confirm it arrives, then move the rest in two or three tranches rather than one. Prefer routes that deliver the native asset rather than a wrapped representation, use the canonical rollup bridge if you can tolerate the delay, and check you are on the correct domain rather than a search-result impostor.

Do I need gas on the destination chain?

Usually yes, and arriving with tokens but no gas to move them is a common and irritating failure. Relay automatically provides destination gas, and several aggregators offer it as an option. If your bridge does not, send a small amount of the destination chain native token first, or plan to acquire it another way.

Cheapest Way to Bridge in 2026

Bridges hold the record for the largest hacks in crypto. The fee difference between them is a few basis points, so read the security model first and the price second.

Published 17 September 2026 · last reviewed 17 September 2026 · figures checked against each protocol's documentation and live quotes on 9 August 2026

For most people the answer is an aggregator: open Jumper, quote your route, take the best one. The rest of this article is why that is the right default, and when it is worth going direct.

The short answer

Best default: Jumper (LI.FI) Quotes Across, Relay, Stargate, Mayan and others live and routes to the winner. You stop needing to know.
Cheapest on L2s: Across 0.04–0.10% on EVM routes, sub-minute. Relayer competition does the work.
Widest chain reach: Relay Reaches chains others do not, and lands you with destination gas already in hand.
Solana routes: deBridge The strongest option when Solana is one end of the trip.
Large native transfers: Stargate Delivers native assets rather than wrapped ones, which is worth a basis point or two by itself.

Why bridging is the riskiest thing you do all week

Moving assets between chains means something has to hold value on one side while releasing it on the other. Every design for doing that has been attacked, and several have been drained for sums in the hundreds of millions. This is not a hypothetical concern raised for balance; it is the single largest category of loss in the history of the industry.

The good news is that the architecture has genuinely improved. There are three generations in use:

Lock and mint. The original design. Assets are locked in a contract on the source chain and a wrapped representation is minted on the destination. The lock contract accumulates enormous value and becomes a target proportional to its size. Most of the famous bridge hacks were this pattern.
Pooled liquidity. Liquidity providers hold assets on both sides and a messaging layer instructs the release. Faster and it delivers native assets, but the pools still hold capital and the messaging layer becomes the thing to attack. Stargate over LayerZero is this model.
Intent-based. You state what you want. A competitive relayer fronts their own funds on the destination chain immediately and is reimbursed from your source-chain deposit afterwards. Very little capital pools anywhere, the relayer takes the timing risk, and competition compresses the fee. Across, Relay and deBridge work this way, and it is the reason both cost and risk have fallen.

Intent-based is structurally better, not perfect. There is still a settlement mechanism to trust, an oracle or validator set behind it, and smart contracts on both ends. Prefer the newer designs and keep the exposure short.

What bridging actually costs

Moving $5,000 of USDC from Ethereum mainnet to Arbitrum.

Across: 0.04–0.10%; $2–$5
Stargate: 0.01–0.06%; $0.50–$3
Relay: quote-dependent; competitive
Aggregator: best of the above; + routing fee
Source gas: Ethereum, varies; $1–$10

On this route the gas frequently costs more than the bridge. That reframes the decision: the difference between the cheapest and most expensive bridge here is a few dollars on $5,000, while choosing to bridge during a congestion spike can cost more than all of them. Timing beats venue selection at retail size.

Below about $500, bridging from Ethereum mainnet is usually not worth doing at all. Withdraw directly to your destination chain from an exchange instead, which is free on Coinbase for native USDC.

The full comparison

Bridges and aggregators compared. Costs are quote-dependent; see the verification date above.
Product	Model	Typical cost	Speed	Chains	Security model	Best for
Across	Intent-based, relayers front the funds	0.04–0.10% on EVM routes	Usually under a minute	Major EVM chains and L2s	UMA optimistic oracle, canonical bridge settlement	USDC between Ethereum L2s
Relay	Intent-based with relayer execution	Competitive, quote-dependent	Seconds on major routes	Widest coverage, including non-EVM	Relayer trust plus escrow	Obscure chains and automatic destination gas
deBridge	Intent-based cross-chain swaps	Tight on ETH and WBTC majors	Seconds	EVM plus Solana	Validator set with slashing	Solana to EVM and back
Jumper (LI.FI)	Aggregator over many bridges	Best-of quoted routes, plus a small routing fee	Depends on the route chosen	30+	Inherits whichever bridge it routes through	Not having to know which bridge is cheapest today
Stargate	Pooled liquidity over LayerZero messaging	0.01–0.06% plus slippage on size	Near instant on supported pools	Broad, LayerZero-connected	LayerZero DVN configuration	Native USDC delivery, larger sizes
Squid	Aggregator built on Axelar	Route-dependent	Minutes on Axelar routes	Broad, including Cosmos	Axelar validator set	Cosmos and other non-EVM ecosystems

Jumper (LI.FI)

An aggregator over most of the others, and the sane default for anyone who bridges occasionally. It quotes Across, Relay, Stargate, Mayan and more in real time and routes to whichever offers the best combination of cost, speed and slippage on your exact pair and size. You stop needing to track which bridge is currently winning which route, which changes with liquidity conditions.

It charges a small routing fee, and that fee is almost always less than the gap between the best and worst available route. The security caveat is inherent to aggregation: your risk is whichever underlying bridge it routes you through, plus LI.FI's own contracts on top. You are adding a thin layer, not removing the underlying one.

Pick it if you bridge occasionally across varied routes and would rather not research the current cheapest option each time.

Skip it if you bridge the same route constantly and already know which venue wins it, where going direct saves the routing fee.

Open Jumper (LI.FI) Jumper (LI.FI) sets its own terms and prices. Check them before signing up.

Across

Typically the cheapest for USDC between Ethereum L2s, settling in the 0.04% to 0.10% range in under a minute. It is intent-based: relayers compete to front your funds on the destination and are reimbursed later, with settlement backed by UMA's optimistic oracle and the canonical bridge underneath.

The competitive relayer market is doing the work, which means pricing is genuinely tight rather than subsidised. Coverage is major EVM chains and L2s, so it is a specialist rather than a universal option. For the single most common bridging task in crypto, moving stablecoins between L2s, it is usually the answer.

Pick it if you are moving stablecoins between Ethereum L2s, which is most bridging, and you want the tightest price.

Skip it if you need a chain outside its coverage, or you are moving an unusual asset.

Open Across Across sets its own terms and prices. Check them before signing up.

Relay

The widest chain coverage here, including places the others simply do not go, and it settles in seconds on major routes. Its most useful feature is unglamorous: it delivers destination gas automatically, so you arrive able to actually do something rather than holding tokens you cannot move.

That combination makes it the best option for small transfers to unfamiliar chains, which is precisely the situation where arriving without gas is most likely and most annoying. For large transfers on well-served routes, Across or Stargate will usually price better.

Pick it if you are going somewhere unusual, or you are moving a small amount and want gas on arrival without a second transaction.

Skip it if you are moving significant size on a major route, where the specialists price tighter.

Open Relay Relay sets its own terms and prices. Check them before signing up.

deBridge

Intent-based cross-chain swaps with the strongest Solana support of anything here. If either end of your trip is Solana, this is usually the right choice, and it produces tight effective costs on ETH and WBTC majors as well. Settlement is in seconds, secured by a validator set with slashing.

On plain EVM-to-EVM stablecoin routes it is competitive rather than dominant, so the reason to reach for it specifically is the Solana leg.

Pick it if Solana is one end of the route, or you are moving ETH or WBTC and want the tightest effective cost.

Skip it if you are moving stablecoins between EVM L2s, where Across typically wins.

Open deBridge deBridge sets its own terms and prices. Check them before signing up.

Stargate

A hybrid: LayerZero messaging plus pooled liquidity, charging roughly 1 to 6 basis points on supported pools plus slippage that grows with size. Its distinguishing property is that it delivers the native asset rather than a wrapped representation, which saves you a swap on arrival and removes a category of confusion about which USDC you are actually holding.

That matters more at size. On a large transfer, receiving native USDC rather than a bridged variant is worth more than the fee difference, because unwinding a wrapped position later has its own cost and its own liquidity risk. The security model rests on the LayerZero DVN configuration, which is worth understanding if you are moving serious amounts.

Pick it if you are moving larger amounts and want the native asset delivered, not a wrapped version to deal with later.

Skip it if you are moving small amounts, where pool slippage and the messaging overhead are proportionally worse.

Open Stargate Stargate sets its own terms and prices. Check them before signing up.

Squid

An aggregator built on Axelar, and the reason to choose it is Cosmos. If your route touches the Cosmos ecosystem, Squid handles it where most of this list does not. On plain EVM routes it is slower and generally more expensive than the intent-based options, because Axelar's validator confirmation takes minutes rather than seconds.

Pick it if your route involves Cosmos or another ecosystem the intent-based bridges do not reach.

Skip it if you are on standard EVM routes, where every other option here is faster and cheaper.

Open Squid Squid sets its own terms and prices. Check them before signing up.

How not to lose the money

More funds are lost to procedure than to protocol failure. Five habits, in rough order of how much grief they save.

Check the domain. Bridge phishing sites buy search ads and clone interfaces exactly. Bookmark the real URL and use the bookmark. Never arrive at a bridge from a search result or a message.
Send a test transaction. Twenty dollars first, confirm it lands, then send the rest. This catches wrong-chain errors, wrong-address errors and interface bugs for the price of a coffee. Do it every time you use a new route, not just the first time ever.
Split large amounts. Two or three tranches rather than one. If something is wrong, you find out on the first tranche.
Check what arrives. Native or wrapped. If your bridge delivers a bridged variant and you expected native, you have a swap to do and possibly a liquidity problem.
Plan for gas. Arriving with tokens and no native gas token is common. Relay handles it automatically; otherwise send a small amount first.

One more, which matters most: do not treat a bridge as storage. Bridge when you need to move, and move on. The exposure window is the risk, and it is entirely under your control. A wallet with transaction simulation, covered in our software wallet comparison, will also catch a surprising share of bridge-adjacent mistakes before you sign them.

What we could not verify

Live quotes. Every cost figure here is a typical range from protocol documentation and observed quotes. Actual cost depends on route, size, liquidity and the moment you ask. Quote before you send.
Relative security. We describe each protocol's stated security model. Comparing the real-world robustness of an optimistic oracle against a validator set against a DVN configuration is not something an article can settle, and reasonable researchers disagree.
Chain coverage. Supported chains change frequently. Check the current list rather than relying on the summary above.

Choosing, in one paragraph

Use Jumper and take the best quote, unless you have a reason not to. Go direct to Across for stablecoins between L2s, deBridge if Solana is involved, Stargate for large native transfers, Relay for obscure chains or when you want gas on arrival, and Squid only for Cosmos. Below roughly $500, skip bridging from mainnet entirely and withdraw straight to your destination chain from an exchange. Whatever you pick, send a test transaction first.

Once the capital lands, see what it actually does:

Open LabelYX: it's free Read-only analytics for any Hyperliquid address. No sign-up, no wallet connection.

Frequently asked

What is the cheapest bridge?

Route-dependent. Across for USDC between L2s at 0.04% to 0.10%, Stargate at 0.01% to 0.06% for native delivery, deBridge and Across tightest on ETH and WBTC. Quote an aggregator on your exact pair rather than memorising this.

Are bridges safe?

They hold the record for the largest exploits in crypto. Intent-based designs pool far less capital and are structurally safer than lock-and-mint, but safer is not safe. Keep the exposure short.

Bridge or aggregator?

Aggregator, unless you bridge one route constantly. The routing fee is usually smaller than the gap between the best and worst route on the day.

How long does it take?

Seconds to a minute on intent-based bridges over major EVM routes, near-instant on Stargate pools, minutes through Axelar. Canonical rollup bridges take days and are the most trust-minimised.

Safest way to move a large amount?

Test transaction, then two or three tranches. Prefer native delivery over wrapped, check the domain from a bookmark, and consider the canonical bridge if you can wait.

Will I have gas on arrival?

Not always, and it is a common annoyance. Relay provides destination gas automatically and some aggregators offer it as an option. Otherwise send a small amount of the native token first.