HIP-3 deployer slashing is a stake-weighted validator vote on the HYPE a builder posted to run one perp dex. After that dex deploys, the spec's official maintain floor is 183 days. A slash burns those coins. Traders in that market keep whatever mark, margin, and the oracle already produced.
The 183-day floor counts from dex deploy. Your first fill is a later event on that same calendar. HaltTrading that settles positions to mark leaves the deploy date in place. A 30-day-after-halt waiting period is a different calendar. Posted listing stake stays slashable through the 7-day unstaking queue.
Permissionless listing and a slashable listing stake sit on the same product.
What HIP-3 deployer slashing is
HIP-3 slashing is the penalty on listing collateral. A qualifying deployer posts HYPE to list one independently margined perp dex. Validators can vote to slash that posting when market operation jeopardizes protocol correctness, uptime, or performance. The coins that come out of a successful vote are burned.
HIP-3 builder-deployed perps covers the product those coins list. The trader-facing residual is what the vote sizes, what HaltTrading does to your book, and what remains in your account after a burn. The project's own HIP-3 article states the trigger in one sentence: "in the event of malicious operation, validators have the authority to slash the deployer's stake by stake-weighted vote" (source: Hyperliquid, HIP-3). The same article's guiding principle is protocol correctness, uptime, and performance. Malicious operation is the trigger language. The size of the burn is a later median of validator votes. A 40 percent drawdown on your silver perp is inventory and margin on that independent book, a separate number from the vote.
HIP-3 listing collateral is one kind of risk among others on whether Hyperliquid is safe. A slash on one builder dex is a local listing event. Venue-wide insurance would be a different product.
The current official mainnet listing stake is 500,000 HYPE, and the same spec says that figure is expected to decrease as infrastructure matures. Extra above the latest requirement can leave. Required collateral that remains stays slashable, including in the 7-day unstaking queue. How posting those coins binds a deployer is the job of stake to deploy.
Suppose the dex deploys, and 60 days later HaltTrading settles every position to mark. The 183-day maintain floor still started at deploy. Day 183 is still counted from that deploy date, not from the settle, and a 30-day-after-halt count would be a different calendar.
How validators size a slash
HIP-3 slashing is sized by a stake-weighted median of validator votes. Trader PnL on that builder book is a different input. Guidelines on the spec run up to 100 percent for invalid state transitions or prolonged downtime, 50 percent for brief downtime, and 20 percent for performance degradation.
"The amount slashed in a given instance is ultimately a stake-weighted median of validator votes". Guideline bands sit on that median. Irregular inputs that cause invalid state transitions or prolonged network downtime can be slashed up to 100 percent. Irregular inputs causing brief network downtime can be partially slashed up to 50 percent. Invalid inputs that cause network degradation or performance issues can be partially slashed up to 20 percent. Quote-asset failure is a different vote, and HIP-3 deployers sit outside slashing related to quote assets (source: HIP-3: Builder-deployed perpetuals).
| Guideline | What the spec names | What the vote sizes |
|---|---|---|
| Up to 100% | Invalid state or long downtime | The full listing stake at risk |
| Up to 50% | Brief network downtime | A partial slash |
| Up to 20% | Degradation or performance | A smaller partial slash |
Those bands size a burn of listing collateral. Your isolated-margin PnL is a separate number, so a 20 percent median and a 100 percent median can both leave your mark where HaltTrading already put it.
The 500,000 HYPE mainnet figure is listing collateral, sized to the listing requirement. A trader's notional is a separate number, so extra above the latest requirement can leave while the required posting stays slashable, including in the 7-day unstaking queue.
What HaltTrading does to open positions
HaltTrading is the operator action that cancels resting orders on that HIP-3 asset and settles open positions to the current mark. Resume can recycle the same asset without a fresh auction. Neither step pays you out of the deployer listing stake.
Suppose you are long a builder-deployed silver perp on that dex, isolated, with resting orders working, when the deployer settles the asset. Cancels hit first. Positions go to whatever mark that dex shows at that moment. If that mark is ugly relative to your entry, the loss is a HyperCore settlement on that independent book. A later validator vote that burns a slice of listing stake leaves the mark you already received in place. Resume can list a dated contract again on the same asset slot, while the 183-day maintain floor still started at dex deploy. A later reading where externalPerpPx moves more than 50 percent versus the start-of-day price is a validator review for manipulation.
Independent margining is why this settlement stays local. How HyperCore works is the onchain book those HIP-3 perps inherit, so matching and margin live on HyperCore. A halt on the builder dex leaves validator-operated BTC running, and it leaves a BloFin HYPEUSDT position untouched. You can be settled to silver mark and still be untouched on another book, which is the point of one listing stake funding one independently margined dex. Enabling cross margin on an asset is irreversible, and users take greater risk using cross margin across dexes with different deployers.
A generic primer on perpetuals versus futures still helps if mark and funding are new. Slash burns listing collateral. HaltTrading settles inventory. Mixing those two verbs is how a trader waits for a rebate after settlement already finalized against mark.
If you needed the book to keep quoting, HaltTrading already ended that quote. A later burn of deployer HYPE is a separate listing event on a separate calendar.
How malice, incompetence, and a stolen key are treated
The spec cares about the effect of deployer inputs on the protocol. A well-meant listing with a bad contract spec can slash, and a compromised key can slash too. Validators vote on that effect. You wait on no finding of intent before your mark moves.
Attempted malicious inputs that skip a protocol issue can still slash. Inputs that do cause protocol issues but that are regular sit outside this slash. Bugs under normal operation that are unrelated to deployer inputs sit outside it as well. Those reports belong on the bounty. On mainnet, any bug that would cause an outage or logical error on nodes or API servers is in scope. Rewards are paid in USDC on Hyperliquid for responsible disclosure (source: Hyperliquid Docs, bug bounty program). A bounty payment is USDC for that disclosure. It is a separate pot from your mark-to-mark settlement and from the listing stake.
Smart contract wallet risks is the generic reminder that an operator key is a control surface, and on HIP-3 that surface can produce irregular inputs validators later vote on.
Subjective oracle-feel disputes are the other edge. The spec says the protocol should stay out of subjective matters, which is why burned coins stay burned instead of being distributed to affected users. Downstream apps and social layers can still react.
Validator jail vs HIP-3 slashing
Delegated staking can jail a slow validator and stop rewards while principal stays in the staking account. Automatic slashing is currently off at that stake layer. HIP-3 slashing is a separate stake-weighted vote on deployer listing stake. The English word slash is reused. The events differ.
How HYPE staking works times the path: instant spot-to-staking, a 1-day delegation lockup, and a 7-day path back to spot. A jailed operator drops out of consensus until it fixes latency. A validator may unjail itself by diagnosing and fixing the causes, subject to onchain unjailing rate limits (source: Hyperliquid Docs, staking). Unjail is a liveness recovery. It leaves HIP-3 listing collateral in place, and it leaves your silver perp on its own mark.
How Ethereum validators work is the other chain's version of an operator bond. Ethereum burns ether for slashable consensus faults and then removes the validator over a long exit, up to 1 ETH burned, with removal taking 36 days (source: ethereum.org, Proof-of-stake rewards and penalties). Hyperliquid's stake layer treats jail as a rewards stop. HIP-3 then uses slash for a builder-dex vote.
Your 400 HYPE delegated to a consensus operator can sit through a jail vote and still be 400 coins in a staking account. Your silver perp settles on its own book. A BloFin instrument sits in neither event.
What LST operators must disclose
Slashing conditions are independent of who supplied the staked HYPE. An LST that funded a HIP-3 deployer carries the burn if that deployer is slashed. Diligence and user disclosure sit on that operator.
A self-bonding requirement is named as something that could make sense. It is a possible design, and that sentence is no live mainnet coupon. FalconX, writing a market note instead of a protocol spec, treats pooled HYPE as a plausible way to source the listing requirement. That note even sketches LST or pooling markets for builders who need HYPE to launch (source: FalconX, The Transformational Potential of Hyperliquid's HIP-3). That is an outsider's product thesis. Diligence and disclosure sit on the operator if your coins funded the listing stake.
If you are the trader, the LST's loss is a separate residual from your mark-to-mark settlement. If your HYPE sits inside the LST that posted the listing stake, you are on the deployer side of the burn. Diligence here is operator due diligence on the listing.
The communicate-to-users clause is the part a pooled token can still fail even when the deployer looks solvent on day one. A later 50 percent median on irregular inputs takes HYPE out of the listing stake by burn, and share-token holders eat that unless the LST's own docs already treated the deployer as slashable listing collateral. Self-bonding, if it ever ships, would change who posts the first slice.
The vote sizes the same way whether the 500,000 arrived as self-stake or as pooled tokens.
What a trader still holds after a slash
After a HIP-3 slash, you still hold whatever the independent book, the oracle that deployer set, and your own margin produced. The burned HYPE is gone from the listing stake. It is no credit on your fill, no BloFin HYPEUSDT rebate, and no jail event on delegated stake.
The current official mainnet listing stake is 500,000 HYPE, expected to decrease as infrastructure matures, so name that integer as today's listing requirement instead of as a dollar headline. The chain, the token, and the other ticker are separate names, covered in what Hyperliquid is. The listing stake raises the cost of listing garbage. CoinGecko's blurb even stretches the venue into those underlyings: Hyperliquid is best known for perpetual futures and spot trading of crypto, equities, commodities, FX, and more (source: CoinGecko, Hyperliquid). A builder silver book can sit in that set and still settle you to mark.
BloFin HYPEUSDT SWAP is live at 75x, listed December 19, 2024 11:30 UTC (source: BloFin instruments API, SWAP). The OpenAPI JSON key is HYPE-USDT. That instrument is a USDT-margined perpetual, and HyperCore deployer collateral is a different posting. The public spot row HYPE/USDT is live too, listed May 30, 2025 13:30 UTC (source: BloFin spot instruments API). HYPERUSDT on the same venue is Hyperlane, a different listing. None of those instruments sits inside the HIP-3 listing stake that validators can burn, so a slash on a builder silver book leaves them in place.
You do not need to memorize the guideline bands. Carry three calendars: 183 days from deploy, HaltTrading to mark on that dex, and a vote that burns listing collateral. If you wanted the product, start with builder-deployed perps. If you wanted what posting the coins binds, that is the listing stake. If you wanted whether the venue as a whole is safe, that is the safety page. After the listing stake burns, your fill is still your fill.
Frequently asked questions
If I only hold an isolated silver perp on that HIP-3 dex, does my fill count as a vote when validators size the burn?
No. The median is stake-weighted among validators, and traders on that independent book are a different set, so your isolated fill is inventory and margin on HyperCore. A 40 percent drawdown on silver adds no weight to the median, and a profitable fill has no veto. A say in that vote needs validator stake, which is a different posting from a builder perp. The listing stake still burns or stays by that median, whether or not you were in the book.
If the mainnet requirement later falls below 500k HYPE, do open traders receive the surplus the deployer can unstake?
No. Extra above the latest requirement can leave as a deployer unstake, and that surplus is no rebate window for open notional. The spec says the requirement is expected to decrease as infrastructure matures, which is a listing-parameter change. Required collateral that remains stays slashable, including in the 7-day queue. A lower published requirement writes you no check for marks you already took.
If I withdraw HYPE from a BloFin HYPEUSDT position onto HyperCore spot, does that start the 183-day HIP-3 calendar?
No. The 183-day floor starts when that perp dex is deployed with listing collateral posted. A withdrawal can put HYPE in your spot account without deploying a dex, posting the 500,000 listing stake, or placing those coins on a slash list unless you later stake them as deployer collateral. HYPERUSDT on the same board is still Hyperlane. Sitting in spot after a BloFin fill is no listing bond.
If a bug bounty pays USDC for a related disclosure, does that reverse a HaltTrading mark I already received?
No. Bounty rewards are USDC paid for responsible disclosure of in-scope bugs, and that rail is a separate unwind from mark-to-mark settlement. HaltTrading already settled you to whatever mark the dex showed. A later USDC bounty to a researcher credits none of your isolated silver, restores no canceled rests, and refills none of the deployer listing stake. Those are three different pots: disclosure, inventory, listing collateral.
Does a 100 percent median on this dex automatically open a replacement book under a new listing stake?
No. A full burn takes listing collateral. The spec documents no auto-relist that opens a fresh independently margined book under a new 500,000 for the same traders. Resume can recycle an asset slot on an existing dex without a fresh auction, which is inventory returning, a different event from a new listing bond. If the listing stake is gone, wait on no protocol coupon that puts 500,000 back and restores your silver size.
If I am cross-margined across two HIP-3 dexes with different deployers, does a slash on dex A seize the margin that was supporting dex B?
No. A HIP-3 slash burns listing collateral on the deployer that got the median. The spec documents no side effect that pulls your cross margin from a second independently margined dex as part of that burn. Users take greater risk using cross margin across dexes with different deployers, a portfolio-risk warning, separate from a shared-stake seizure. Enabling cross margin on an asset is irreversible. Isolated silver on A can settle to mark without rewriting B, and a later burn on A's listing stake is no rebate on B.
If my 400 HYPE is only delegated to a consensus validator, does a HIP-3 deployer slash that validator voted on burn those 400 coins?
No. Delegated HYPE sitting with a consensus operator is HIP-3 listing collateral only when those coins were the posted stake, so a validator's stake-weighted vote can size a burn on a deployer while leaving your 400 in a staking account. Jail on that operator can stop rewards while principal stays. The silver perp still settles on its own book, and the 400 coins stay a staking balance until you unstake them on that 7-day queue.
Researched and written by the BloFin Academy editorial team with AI-assisted drafting. Updated August 2026. Primary sources include the Hyperliquid HIP-3 spec, the HIP-3 article, staking docs, the bug bounty program, and BloFin's public instrument API. Protocol and listing facts independently verified against cited sources current as of August 2026.
This article is educational and general in nature, not financial or investment advice. Cryptocurrencies like HYPE carry real risks, including price volatility, deployer and oracle risk on builder markets, liquidation of borrowed size, venue risk, and the chance of losing funds. Nothing here is a recommendation to buy, sell, hold, deploy, or trade any HIP-3 market. Do your own research, and consider speaking with a licensed professional before making financial decisions. BloFin does not provide investment advice.