Let’s Talk Cross-Chain w/ The Across Protocol

By LI.FI | LiFi | 20 Jun 2022

Let’s Talk Cross-Chain is a series hosted by LI.FI where we interview founders and teams building key infrastructure for the multi-chain future.

For this interview, we’re joined by the Across team. Across is an optimistic cross-chain bridge protocol that allows users to execute transactions between chains nearly instantaneously. They accomplish this by using an optimistic oracle, bonded relayers, and single-sided liquidity pools.

Let’s dive in!



Welcome to the LI.FI’s Let’s Talk Cross-Chain Interview Series. First, please introduce yourself, as well as the origins and development history of Across.


I’m Korpi and I am the product strategist at Across. My role involves helping Across win. This means the highest volume, fastest, cheapest, and most secure solution for cross-chain transfers. I spend most of my time trying to make the product better and working with integrations like LI.FI to get more users for Across.


I’m Nick Pai and I am the technical lead at Across. My role involves writing code: the smart contracts and the off-chain “bots” that interact with them, and managing our full-stack engineering team. Now that the code is deployed and the product has launched, I spend most of my time considering edge cases where our system may fail, simulating and patching these issues, and also prodding and poking the system in prod. My primary goal at the moment is to make the system more robust.

Our history: In June of 2021, the discussion around building a cross-chain bridge began on our Risk Labs team, the team behind UMA and Across. We were looking to develop a product that would showcase the power of UMA’s optimistic oracle. During this discussion, the topic turned to L2s and bridging. The optimistic oracle seemed like it would be a great fit to secure cross-chain transfers. With this in mind, we began building what we called “the fast and fully insured bridge” with 3 key features in mind — fast, fully insured, and capital efficient. The team did an incredible job refining everything (for example, bringing costs down for relayers, from 1.4mm units of gas to about 200K) within a short period of time. Across was conceptualized in June and launched in November of that same year. Within the first week of our launch, we crossed $10.4 million in total volume. We’re now at over $250 million across the bridge and we just launched V2 of Across this past week! If you would like to learn more about our history, please read our Getting Across series on Medium.

Can you go into a bit of detail on how Across leverages UMA’s optimistic oracle and what the UMA token’s role is in Across’ design?


You can think of Across as an “insured bridge” where any user who wants to get from chain X to chain Y can purchase insurance that guarantees that the 10 tokens (using 10 as an example) they want to send from X will arrive on Y, less the insurance fee. The across liquidity providers and relayers insure the transfer by crediting the user their ~10 tokens on chain Y, and receive the insurance fee in exchange. The optimistic oracle is used to settle insurance claims, which are submitted when someone feels that a user was not correctly paid on chain Y: perhaps they were paid the wrong token on Y, or the wrong amount, or credited to the wrong address. In more than 99% of cases, insurance claims will never be submitted, which is great because users get fast bridged transfers and relayers and LPs earn insurance fees. The UMA token is held by voters who settle the insurance claim, so you could say Across is secured by the UMA token holders.

Can you talk more about why you decided to go the optimistic rollup route? Why are you so Optimistic about Optimistic solutions :)?


The core value of an insured bridge is that the user will always get their expected amount of tokens on chain Y, but the design does not guarantee how quickly it will arrive. This is different from a cross-chain AMM like Thorchain or Hop, where the time to wait is guaranteed but the amount is different. Unlike AMM-based bridges, insured bridges do not rely on price arbitrage and thus have no slippage. Furthermore, it’s easier to convert an insured bridge system into an insured messaging system, where anyone can send an arbitrary message across networks that have insured integrity. When I say integrity, I mean that the message is unaltered as it is relayed across networks.

How big of a role does decentralization play in Across’ design? In what areas was decentralization prioritized and where were certain centralized trade-offs necessary, specifically in the context of the three main players in Across’ design: relayers, LPs, and the Optimistic Oracle?


Across was built with decentralization at its forefront. The trusted actors in the Across system are the Across governance, currently a multisig and extremely soon to be converted into a DAO secured by the UMA DVM, the UMA DVM that settles insurance claims, and the governance system of the L2s that we support: Optimism, Arbitrum, Polygon, and Boba. The L2 governance that controls the canonical bridges from Mainnet to the L2s has the ability to modify messages sent between Mainnet and the L2s, so Across is reliant on that. There are no extra trusted agents, which is why Across doesn’t currently support Avalanche which doesn’t have a canonical bridge to Mainnet. Deploying contracts on these L2s already means that we are assuming the risks that come with their governance systems, so we don’t want to add any other agents to the mix.

Can you talk a bit more about the V2 release? What’s new? What’s the same? And what still needs to be changed?


Yes! We’re incredibly excited about V2’s release. There are a few new features you can look for in Across’ V2.

We now feature L2 to L2 transfers as well as an additional chain — Polygon. Across V2 also prides itself on its capital efficiency. There are a few ways we do this. Across V2 enables L1->L2 and L2->L2 transfers through the use of a hub and spoke model. The hub pool sits on L1 (similar to Across V1) but there are now spoke pools on all of the L2s. These spoke pools are used to refund relayers on those L2s when L1->L2 or L2->L2 transfers happen. V2 also now batches and nets repayments in order to save on transfers and hence makes the operation more efficient.

With our new version, you can still expect the same pillars that we pride ourselves on to be present — security, speed and cost-efficiency.

Speaking of V2 and all the exciting things going on at Across, we have to ask 1) wen token and 2) how are the plans for passing off the governance of Across from UMA leaders to the community going? We would love an update on the “fair fair launch” :).


Token is in the works; we’re aiming for sometime in Q3. We’re working on bringing together all of the distribution components and building out the interface for the rewards locking mechanism.

In regards to passing off the governance of Across, we’re hoping to introduce a governance mechanism at the time of token launch that will put the future of the protocol into the hands of ACX token holders. Risk Labs will continue to tend to the admin duties of the protocol until this process has been automated.

What are Across’ plans for implementing new chains? Is it just sticking to Ethereum and Optimistic L2s for now?


We are constantly evaluating new chains to add. Some of the chains lack a canonical messaging bridge so will be more difficult to add. We are exploring various options here to make this possible, such as building our own arbitrary messaging bridge or using an existing one that is decentralized.

What are the biggest challenges you’re facing currently?


Currently, we are looking to promote more awareness around Across. We have the best fees for large transfers and want the world to know about it! Our average bridge user is significantly bigger than our competitors’ average user, but we also want the smaller users too. We continue to work to make sure our fees are extremely competitive and that we provide a great user experience.

What are you most looking forward to launching or attempting to solve in the next few years when it comes to bridging infrastructure at Across?


Solving the arbitrary messaging system would be nice as that can enable us to add more chains to Across. Imagine that there was a way to send messages from any chain to any other chain and have those messages secured by the UMA DVM. Using this hypothetical system, we could support Across from any chain to any other chain while only trusting the governance systems of those chains + the UMA DVM.

I’m also excited to help other teams build cross-chain dapps that use the UMA optimistic oracle too. We originally built Across to demonstrate the power of the UMA optimistic oracle to secure cross-chain messaging. It just so happened that insured token bridging was the most obvious and simplest way to take advantage of cross-chain messaging. Across eventually evolved into a standalone product.

What are your views on bridge aggregation, ie. LI.FI?


We think bridge aggregation is the future and most bridging activity will migrate to aggregators. This is one of the main reasons why we put such emphasis on keeping our fees low.

Who are the investors who have already invested in your enterprise, and what info can you provide about them?


None at this point. Across was built by Risk Labs which is the foundation behind UMA Protocol.


Get started With LI.FI Today:

To learn more about us,

or try our any-2-any swaps NOW at transferto

How do you rate this article?




The ultimate cross-chain liquidity aggregator, aggregating cross-chain liquidity networks to DEXs, calculating you the best cross-chain swaps. The future is cross-chain and we make sure you don't have to care. #DeFi


The ultimate cross-chain liquidity aggregator, aggregating cross-chain liquidity networks to DEXs, calculating you the best cross-chain swaps. The future is cross-chain and we make sure you don't have to care. #DeFi

Send a $0.01 microtip in crypto to the author, and earn yourself as you read!

20% to author / 80% to me.
We pay the tips from our rewards pool.