Add Across Protocol as recommended bridge choice

GMX should add Across Protocol (@AcrossProtocol) as a recommended fast bridge to send ETH to Arbitrum. Right now I see fast bridges recommended on the “Sending Tokens” section of github.

GMX can earn referral rewards for sending volume to Across thru this link. They can also pre-populate the link with desired bridging chains/assets as per below:"ETH"&referrer=GMX_Wallet
GMX_Wallet = Ethereum Wallet Address for GMX to accumulate $ACX rewards

Across is the fastest, cheapest and most secure cross-chain bridge for assets between Ethereum and Arbitrum.


Is there actual evidence to these claims… ?

I don’t say that just to be facetious. Many users regularly use bridges, and there’s a multitude of them - and more new ones every week. I, personally, wouldn’t mind not having to endlessly compare Multichain, Synapse,, Hop, Connext and the rest of them…

And there are very significant sums on the line.

So, if you want to convince the community, do share the data that makes us prefer Across.


hi @Jonezee - thank you for the feedback/comments and totally valid. Fast/cheap/secure are our main tenants, I’ll give a bit more context about each below.

Fast - our transfers are typically 1-4 mins, you can see this from the messaging in our UI, or if you try us out. From using all of our competitors, this is in the top tier of fastest bridges.

Cheap - when using our bridge (+ the referral program), our bridge is (almost always) the cheapest option. With the referral program, you can get up to 80% of the bridge fee rebated back (this is currently 160% since we are applying a 2X boost). I won’t flood the thread with screenshots, but if you want to give me your email, I’m happy to send you as many as you’re willing to view showing examples of our fees vs competitors. The reason we are able to be low cost is Across bridges the native tokens over the canonical bridges with relayers/LP to provide loans of native tokens while assets are in transit. There is no AMM style swapping where slippage is incurred.

Secure - this is hard to quantify. We are an optimistically verified bridge and we believe that model is secure. There are some good blog posts about “the spectrum of trust” (written by Li.Fi) in bridges if you want to get other opinions about various bridge architectures and their relative security.

One last comment, we are a fan of aggregators! We are integrated with most of them already and we do well as our prices are competitive. The disadvantage with GMX going that route is GMX would not earn the referral rewards if they send users thru an aggregator to get to Across. Having said that, I understand why an aggregator could be appealing for GMX and we would be in favor of that route too.

By the way, I think GMX referral program is killer and it inspired some of the aspects of the Across referral program (the only bridge that has one that I’m aware of). We want to build a leaderboard similar to the ones your community has done.

You can read about the Across referral program here: Referral Program - Across


I prefer the synapse protocol, i really like the layout and it’s a profissional job

But to have a huge aount of bridges and partners is better because it increases the popularity of gmx

Thank you for proposing this.

We are engaged with multiple bridge and bridge aggregators on a non-exclusive basis to better support GMX including bridge + transaction combinations that make buying and staking GMX / GLP or having collateral(+gas) on chain easier for existing new users joining the GMX ecosystem.

If Across can help with this, feel free to DM @coinflipcanada on telegram or twitter and happy to have a more robust technical and partnership discussion.


Would recommend any type of explicit bridge integration be run through this (or similar) type of risk framing methodology: L2Bridge Risk Framework - #28 by randomishwalk - Methodology & Framework - L2BEAT

1 Like

thanks for all the engagement! I have DMd @coinflipcanda . is there anything else we can do to move this along?

Am not a GMX contributor, but out of curiosity and for my personal ease of mind, is there a simple way you can lay out the key security and trust assumptions that underly Across Protocol? I have not yet dug into it myself tbh, so would appreciate if you can lay it out here simply for all of us to assess :innocent:

Specifically the Across security model and double-clicking on @Jonezee’s good questions:

  • Upgrade privileges and admin keys? Do they exist? And if so, what and how governs access to these?
  • Optimistic bridge security – how many entities serve as “watchers” today? Do you have working fraud proofs and has there ever been an example of a fraud proof that has worked as expected on mainnet?
  • Fraud proof mechanism – what happens when a successful challenge is made? What is the bridge rollback mechanism? Conversely, a fraud proof system means there is some time to finality (depending on which POV you take in terms of the various counterparties involved in a bridge transaction) – what is that latency for Across?
  • Can you please link us to the latest audit, just so that all the relevant information is here for folks to assess?

thank you for the questions. here are the answers.

  1. Yes, there is an admin multisig comprised entirely of hardware wallets of known/doxxed members of the Risk Labs leadership team. This will be decentralized once a token is released.

  2. Risk Labs runs a watcher. We cannot be sure how many others run watchers, but there is a reward for correctly disputing an invalid proposal. There have been multiple disputes on mainnet that have been correctly adjudicated by the UMA Optimistic Oracle system.

  3. A successful challenge throws out the proposed bundle entirely. There is a delay, but end-users don’t see that delay because relayers front the capital to the user and then are repaid by the bundle after the fact if their relay was valid.

  4. Audit report is here: UMA Across V2 Audit - OpenZeppelin blog

1 Like

Awesome, appreciate the thorough answer!

1 Like