Hyperliquid staking is HYPE delegation with a 7-day unstaking queue

Hyperliquid staking is HYPE delegation to HyperCore validators, with rewards automatically compounded and a fixed two-step exit. A fresh delegation remains locked to its validator for 1 day. After that, undelegation returns HYPE immediately to the staking balance, but only a separate staking-to-Spot transfer starts the 7-day unstaking queue. Validator commission, uptime, the network-wide reward curve, and the five pending-withdrawal limit determine whether the arrangement fits your liquidity plan.

Updated on 6 Aug 2026

The short version: An address may hold only five pending staking-to-Spot withdrawals at once, so batching preserves queue capacity.

Undelegating alone leaves HYPE outside the 7-day exit queue

The HyperCore exit path has two separate actions: undelegate from a validator, then transfer from Staking to Spot. Undelegation becomes available after the delegation's 1-day lock and returns the selected amount to the staking balance immediately. That balance is no longer assigned to the validator, yet it has not entered the withdrawal queue and is not available in Spot.

The fixed 7-day clock begins only when the staking-to-Spot transfer is signed. Each address may hold no more than five pending withdrawals, so several small transfers can consume every slot before a larger exit is requested. If HYPE still appears as an undelegated staking balance after seven days, the queue did not run late; the second action was never initiated. Queue completion places the amount in the HyperCore Spot balance.

Partial and full undelegations follow the same two-step path. Moving only part of a position leaves the remainder assigned to its validator and eligible for rewards, while the selected amount becomes available for a queue request. The five-slot ceiling applies to the address, not separately to each validator used by that address.

Validator commission reduces the reward credited to a delegator

Validator commission is the explicit reward cost in Hyperliquid staking. HyperCore assigns gross rewards in proportion to delegated stake, then the chosen validator's commission reduces the amount credited to delegators. The protocol accepts a commission increase only when the new rate is 1% or less. That restriction governs increases; it should not be read as a universal 1% ceiling on every displayed commission.

Worked example with hypothetical changing inputs: assume 2,000 HYPE delegated, a hypothetical 3% gross annual reward rate, a hypothetical 5% validator commission, and a hypothetical 365-day holding period with uninterrupted reward production. The net nominal rate is 3% × 95% = 2.85%. Applying daily compounding gives 2,000 × [(1 + 0.0285 ÷ 365) ³⁶⁵ − 1] = 57.82 HYPE, producing a hypothetical ending balance of 2,057.82 HYPE. The protocol's live rate moves with total network stake, so 3% is an illustration rather than a forecast.

More network stake lowers the HYPE reward rate

The HYPE reward curve is inversely proportional to the square root of total HYPE staked across the network. More aggregate delegation therefore lowers the annual rate, while less aggregate delegation raises it. Rewards come from the future emissions reserve rather than validator commission, and the displayed annual rate should be treated as a network output rather than a promised return.

At 400 million HYPE delegated network-wide, the published curve yields an approximate annual reward rate of 2.37% before validator commission. Rewards accrue every minute, are distributed once each day, and are automatically redelegated to the same validator. HyperCore calculates them from the delegator's minimum balance during each staking epoch. One epoch spans 100,000 consensus rounds, approximately 90 minutes on mainnet, so a mid-epoch addition does not retroactively raise the minimum for the elapsed portion.

Validator thresholds determine valid delegation destinations

Validator eligibility shapes the set of usable delegation destinations. Hyperliquid's active set consists of the top 24 validators by stake, while an individual validator needs 10,000 HYPE of self-delegation to become active. That self-delegation is locked for 1 year. It is the validator's commitment, not a minimum imposed on ordinary delegators.

If self-delegation falls below 10,000 HYPE, the validator enters undelegate-only mode and stops accepting fresh delegation. Existing delegators can still reduce their positions. A holder may split HYPE among any number of validators, and each new validator allocation receives its own 1-day lock. Splitting changes exposure to commission and reward interruptions, but every later staking-to-Spot transfer still uses the same address-level 7-day queue and five pending-withdrawal cap.

The validator set and consensus stake remain fixed within each 100,000-round staking epoch, then update at an epoch boundary. A rank change therefore does not rewrite a completed reward interval halfway through it. Validator selection should account for active-set position, commission, and operating continuity, since delegated stake influences both block production and the delegator's reward stream.

Jailing pauses validator rewards until consensus participation resumes

Validator jailing stops reward production for that validator's delegators. HyperBFT validators submit jail votes when a peer fails to respond with adequate latency or frequency; a quorum requires more than two-thirds of total stake. A jailed validator leaves consensus and produces no rewards until onchain unjailing succeeds. Jailing does not move delegated HYPE, so the holder must wait or undelegate after the 1-day lock. Regular delegation currently has no automatic slashing, making a reward pause distinct from a protocol deduction from delegated principal.

HyperEVM balances must move to HyperCore before delegation

HYPE held on HyperEVM cannot be delegated directly because native staking lives in HyperCore. The holder first transfers HYPE from HyperEVM to the HyperCore Spot balance, moves it from Spot to Staking, and then selects a validator. The Spot-to-Staking leg is instant; the 1-day validator lock begins with delegation, not with the earlier cross-component transfer.

HyperEVM uses chain ID 999, represents native HYPE with 18 decimals, and follows Ethereum's EIP-1559 gas model. HyperCore's HYPE token specification uses 8 wei decimals and 2 spot size decimals; those spot-order settings do not reduce the precision of staking accounting. A HyperEVM-to-HyperCore transfer consumes HYPE gas on the EVM side. Both components share HyperBFT consensus, but their balances remain operationally separate until the transfer is completed.

Six staked-HYPE thresholds reduce trading fees

Staked HYPE also sets a trading-fee discount tier, separate from delegation rewards. The Wood tier begins above 10 HYPE and discounts eligible trading fees by 5%; Bronze begins above 100 HYPE for 10%; and Silver begins above 1,000 HYPE for 15%. The thresholds are strict "above" tests, so a balance exactly equal to a boundary has not crossed it.

The upper tiers continue the same schedule: Gold starts above 10,000 HYPE for 20%, Platinum above 100,000 HYPE for 30%, and Diamond above 500,000 HYPE for 40%. Trading and staking from the same address require no linking action. A separate trading address can be linked to a staking address, but the link is permanent, the staking address gains unilateral control over the trading address, and the staking address stops receiving its own staking-related fee discount after linking.

Different interfaces still write to one HyperCore staking state

Hyperliquid, Nansen, Validao, and HypurrScan present different views of the same HyperCore delegation state. Choosing another interface does not create a separate staking product, restart a completed lock, or shorten a pending withdrawal. The validator address, delegated amount, lock timestamp, rewards, and queued transfers remain attached to the user's blockchain address.

MetaMask, Rabby, WalletConnect, and Coinbase Wallet provide signing paths for an EVM-style address; changing wallet software while retaining that same address leaves the staking record unchanged. Developers can read the same state through the delegation, delegator summary, delegator history, and delegator rewards queries. Validator addresses use a 42-character hexadecimal format, which gives an exact identifier when display names or interface ordering differ. Comparing that address across views avoids treating a presentation change as a delegation change.

An eight-day reversal test identifies suitable staking capital

A newly delegated HYPE balance needs at least 8 days to reach Spot when the decision is reversed immediately: 1 day for the validator lock, followed by the fixed 7-day staking-to-Spot queue. HYPE that has already cleared its validator lock still needs seven days after the transfer is initiated. This timing is the clearest fit test because neither validator quality nor a higher reward rate removes the exit delay.

At a protocol level, Hyperliquid staking fits HYPE intended to remain delegated beyond a near-term liquidity window. Keeping a planned trading amount in HyperCore Spot preserves immediate Spot availability, while retaining HYPE on HyperEVM preserves access to EVM applications and gas. Delegated HYPE serves neither role while assigned to a validator. A staged allocation can separate long-horizon delegation from balances reserved for a known transfer, purchase, or trading date.

Queue-aware batching preserves all five withdrawal slots

The unstaking queue permits five pending staking-to-Spot transfers per address, and every accepted transfer waits exactly 7 days. Batching amounts that share the same target date preserves queue capacity. Staggering transfers makes sense only when their intended availability dates differ, because a sixth request must wait until one of the existing five completes.

A precise exit record includes the validator lock expiry, the undelegated amount, the staking-to-Spot initiation timestamp, and the expected completion timestamp seven days later. Rewards stop accruing on the amount once it is no longer delegated, even while that HYPE remains in the staking balance or withdrawal queue. Pending entries complete independently, and each completion opens one queue slot. Completion returns HYPE to Spot; moving it onward to HyperEVM is a separate transfer with its own gas requirement.