If you opened a Hyperliquid account without entering a referral code, you left money on the table. Not metaphorically โ literally. The platform offers a 4% fee discount on every trade until you cross the $25M lifetime volume cap, and that discount can only be activated by typing a referral code on signup. There is no way to backdate it after the fact.
Most traders who learn about this discount reach for the same mental shortcut: "4% of a 0.045% fee is basically nothing." The arithmetic is technically correct but the conclusion is wrong, because the cap is huge. $25M of cumulative perp volume is what an active trader does over a year or two, and the rebate stacks with HYPE staking discounts and with the volume-tier ladder. Compounded across all three, you can land in a fee regime that is structurally cheaper than what you would pay on most centralized exchanges.
This guide walks through exactly how the 4% discount works, what your savings look like at different volume profiles, how it stacks with the rest of the Hyperliquid fee stack, and what to do if you already signed up without it.
> Note on methodology and scope: I rebuilt every number below from the official Hyperliquid fee schedule on 2026-04-28 using a spreadsheet, after an earlier draft of this article shipped with an arithmetic error in the worked example (a 100 HYPE stake mis-classified as Bronze when the docs require *strictly greater than* 100 HYPE for that tier). I have not personally exhausted the $25M referral cap โ this is fee-stack analysis, not lived trade testimony. Every effective rate is computed inline so you can paste it into your own sheet and verify against the current fee page.
What the 4% discount actually is
Hyperliquid's referral program is documented in the official docs (Hyperliquid referrals). The relevant clause:
> "Using a referral code will give you a 4% discount on your fees for your first $25M in volume."
Three things to notice in that sentence:
1. It is a discount on fees, not on notional. A 4% discount on a 0.045% taker fee means your effective rate becomes 0.0432%, not 0.045% minus 4%. People misread this constantly.
2. The cap is per-account, in cumulative volume. It is not a 30-day window or a calendar reset. Once you have traded $25M, the discount expires for that wallet. 3. It must be applied at signup. Hyperliquid does not let an existing user attach a referral code retroactively. The slot is single-use and time-locked to account creation.The referrer (the person whose code you used) earns a share of the fees you generate. That is how Hyperliquid funds the program: the protocol does not subsidize your discount, the referrer's earnings line does. From your perspective as the new user, this is a free 4% off the top with no catch beyond having to type a code.
If you are still figuring out how to open an account and bridge USDC in the first place, start with the Hyperliquid getting-started walkthrough before worrying about fee math โ the discount is only useful if you actually trade.
The base fee schedule (as of 2026-04)
You cannot reason about a discount without knowing what it is discounting. Hyperliquid publishes its full fee schedule at Hyperliquid fees. The tier-0 base rates, which is what almost every new account sits on, are:
| Market | Maker fee | Taker fee |
|---|---|---|
| Perps | 0.015% | 0.045% |
| Spot | 0.040% | 0.070% |
Volume tiers kick in based on a rolling 14-day weighted volume. Tier 1 (over $5M of weighted volume) starts trimming the rate, and the ladder runs all the way up to tier 6 (over $7B). Spot volume counts double toward the tier, which matters more than you would think for traders who use Hyperliquid's spot markets to manage HYPE exposure.
For most retail traders the relevant number is the tier-0 perp taker fee, 0.045%. That is the rate the 4% discount is applied to. The math is simple:
0.045% ร (1 โ 0.04) = 0.0432%
That looks like a microscopic difference until you scale it across the full $25M cap.
The $25M cap, in dollars
Here is the number that matters. If you trade exactly $25,000,000 in perp volume as a pure taker before the discount expires:
| Line item | Without discount | With 4% discount |
|---|---|---|
| Effective taker fee | 0.045% | 0.0432% |
| Total fees on $25M volume | $11,250 | $10,800 |
| Savings | โ | $450 |
For a maker-heavy trader (someone who works limit orders into the book), the math shifts:
| Line item | Without discount | With 4% discount |
|---|---|---|
| Effective maker fee | 0.015% | 0.0144% |
| Total fees on $25M volume | $3,750 | $3,600 |
| Savings | โ | $150 |
How long does it actually take to hit the cap?
The $25M cap is more reachable than it sounds because perp volume is *gross notional traded*, not net P&L. A single round-trip on a $10,000 leveraged position eats $20,000 of your cap, and most active traders open dozens of round-trips a week.
A rough table of how long the cap lasts at different activity levels:
| Daily volume | Days to $25M | Months to $25M |
|---|---|---|
| $5,000 | 5,000 | ~164 |
| $25,000 | 1,000 | ~33 |
| $100,000 | 250 | ~8 |
| $500,000 | 50 | ~1.6 |
| $1,000,000 | 25 | <1 month |
The 14-day weighted volume that determines your tier is computed independently of the referral cap, so even after the 4% discount expires you are not stuck at tier-0 rates if your trading has been heavy.
Stacking with HYPE staking discounts
The 4% referral discount is one lever. The HYPE staking discount is the other. The official tiers, as of 2026-04, look like this:
| Tier | HYPE staked | Fee discount |
|---|---|---|
| Wood | > 10 | 5% |
| Bronze | > 100 | 10% |
| Silver | > 1,000 | 15% |
| Gold | > 10,000 | 20% |
| Platinum | > 100,000 | 30% |
| Diamond | > 500,000 | 40% |
The Wood tier is genuinely accessible โ at the time of writing, ten HYPE is a small position, well within reach for any trader who is using the platform regularly. The big jumps (Platinum at 100k HYPE, Diamond at 500k) are aimed at market makers and institutional desks. Read the threshold strictly: "> 100" means more than 100, not equal to 100. A wallet holding exactly 100 HYPE qualifies for Wood (5%), not Bronze (10%). I tripped on this myself in the first version of this analysis and have seen the same off-by-one mistake in third-party explainers โ when you size your stake, leave a buffer above the integer threshold rather than landing exactly on it.
Like what you're reading? Try it yourself โ this link supports ChartedTrader at no cost to you.
Activate the 4% fee discount on Hyperliquid โA staking user and a trading user can be linked, which means you can stake HYPE in one wallet and have the discount applied to a separate trading wallet. This matters for anyone who runs a cold-wallet staking setup and a hot-wallet trading setup.
The combined effect on a tier-0 perp taker who uses the referral code and stakes into the Wood tier:
0.045% ร (1 โ 0.04) ร (1 โ 0.05) = 0.04104%
That is an 8.8% effective discount versus the bare base rate. Push it to Bronze (must be strictly more than 100 HYPE staked, 10% discount) and you land at 0.03888% โ a 13.6% effective discount.
For the full mechanics on how staking discounts and Hyperliquid's gasless settlement model interact, the HYPE staking + gasless trading guide is the deeper read.
Lifetime savings: a worked example
Most "lifetime savings" claims are marketing fluff. To make this honest, I will work through a specific trader profile and show every assumption. I built this in a spreadsheet on 2026-04-28 and verified each cell against the fee docs; if you spot a number that looks off, paste your own numbers in and tell me โ I want this corrected, not defended.
Profile: Active retail trader, mostly perps, 60% taker / 40% maker by volume, $50,000 average daily volume, 250 trading days per year. Stakes 150 HYPE to qualify for Bronze (the 10% tier โ well above the strict ">100" threshold). Uses referral code at signup. Annual gross volume: $50,000 ร 250 = $12,500,000 Year 1 fee math:| Component | Volume | Effective rate | Fees |
|---|---|---|---|
| Taker (60%) | $7.5M | 0.045% ร (1 โ 0.04) ร (1 โ 0.10) = 0.03888% | $2,916.00 |
| Maker (40%) | $5.0M | 0.015% ร (1 โ 0.04) ร (1 โ 0.10) = 0.01296% | $648.00 |
| Total | $12.5M | $3,564.00 |
| Component | Volume | Effective rate | Fees |
|---|---|---|---|
| Taker (60%) | $7.5M | 0.045% ร (1 โ 0.10) = 0.0405% | $3,037.50 |
| Maker (40%) | $5.0M | 0.015% ร (1 โ 0.10) = 0.0135% | $675.00 |
| Total | $12.5M | $3,712.50 |
That is what the discount is worth, in honest dollars, for a representative profile. It is not a fortune, but it is free, and it is reliable.
The bigger compounding lever in the calculation above is the HYPE staking, which runs forever. Bronze tier alone, against the no-discount baseline of $4,125/year (= $12.5M ร the volume-weighted base rate of 0.033%), saves this trader $412.50 per year permanently ($4,125 โ $3,712.50). That is roughly 2.8ร the annual referral savings, and it never expires. Stack them while you can.
What if you already signed up without a referral code?
Hyperliquid does not allow retroactive referral attachment. There is no support ticket that fixes this, no admin panel toggle, no "I forgot, can you apply it now" workflow. The protocol's position is that the slot is sealed at account creation.
There are two practical paths if you are already in this situation:
1. Use a fresh wallet. Each EVM address is a separate Hyperliquid account. You can spin up a new wallet, deposit USDC, sign up with a referral code, and run that wallet through the $25M cap. This works but it splits your trading history and your staking position across addresses, which is operationally annoying.
2. Stake HYPE on the existing account. Staking discounts do not require a referral and they are not capped at $25M. If your discount strategy was 4% from referral plus some staking, doubling down on staking is the cleaner path forward.For most users with an existing unused account, option 2 is the right call. The 4% referral is worth a few hundred dollars across two years; splitting your account across two wallets to chase it costs more in operational friction than it saves.
Common mistakes that destroy the discount
A few patterns that come up repeatedly in trader discussions and that the math above makes obvious:
- Signing up the trading account but not the staking account. If you stake HYPE in a separate cold wallet and never link it to the trading wallet, the staking discount does not apply. The link has to be set explicitly on the staking page.
- Routing all volume through taker orders. The 4% discount is a multiplier on whatever fee you pay. Paying 0.045% ร 0.96 is still meaningfully more than 0.015% ร 0.96. The biggest single fee win on Hyperliquid is making, not the referral.
- Treating referral and staking as alternatives. They stack multiplicatively. There is no reason to pick one.
- Burning the cap on test trades. The cap is volume-counted, not trade-counted. If you spend your first month learning the platform with leveraged practice trades, you can chew through several million dollars of cap without realizing it. Practice with small position sizes early.
- Sizing your stake exactly on a tier boundary. The thresholds are strictly greater-than. 10 HYPE does not qualify for Wood; 100 HYPE does not qualify for Bronze. Round up.
- Using leverage that does not match the strategy. Higher leverage does not change the fee math directly โ fees are charged on notional, not margin. But higher leverage does make your gross volume balloon for the same equity, which accelerates how quickly you exhaust the $25M cap relative to your actual portfolio size.
Where the 4% sits in the full Hyperliquid fee stack
Putting it all together, the levers a Hyperliquid trader has on fees, ranked by how much they typically move the needle:
1. Maker vs taker. A maker order pays 0.015% vs 0.045% taker โ a 67% reduction. This is by far the largest lever.
2. Volume tier. Once you cross the $5M / 14-day threshold, every higher tier shaves another slice. Tier-1 trims taker to 0.040%; the higher tiers get aggressive. 3. HYPE staking. Wood โ Diamond gives 5% โ 40%. Permanent, not capped. 4. Referral discount. 4% on first $25M volume. Capped, one-time activation.In dollar terms, for most traders, the order is roughly: maker routing > HYPE staking > volume tier > referral. But unlike the others, the referral has a hard zero-cost activation: you do not need to stake anything, you do not need to trade more, you just need to type the code on signup.
If your trading account already exists and the answer is "I never typed a code," the lesson is to stop missing the next free thing. Hyperliquid frequently runs no-deposit promotions, partner reward distributions, and staking incentives that all operate on the same principle: opt-in at the right moment or get nothing. Set a calendar reminder to check the Hyperliquid blog and the docs once a month.
Bottom line
The 4% referral discount on Hyperliquid is a small but real lever โ somewhere between $150 and $450 saved per $25M of volume depending on your maker/taker split, and roughly $148.50/year for a representative $12.5M/year profile. It is not the biggest fee saver on the platform, but it is the only one with truly zero cost: no staking lockup, no volume requirement, no behavioral change. You type a code on signup and the rate goes down for two years of average-trader activity.
Stacked with Bronze-tier HYPE staking and disciplined maker order routing, the combined effective rate on tier-0 perps drops from a base of 0.045% taker / 0.015% maker into the 0.039% / 0.013% range. That is the regime where Hyperliquid starts being structurally cheaper than most CEX alternatives, and it is where most serious traders on the platform actually live.
If you have not signed up yet, the only thing standing between you and the discount is a single referral code. Activate the 4% fee discount on Hyperliquid before you make your first trade โ there is no second chance to apply it.
*Disclosure: The Hyperliquid signup link in this article is a referral link. Using it gives you the 4% fee discount on your first $25M of volume; it also gives this site a share of the trading fees you generate. The discount you receive is identical regardless of whose code you use. All fee math was recomputed from the official docs on 2026-04-28 โ if a rate has changed since publication, the formulas in the worked example are still valid; just substitute the current numbers.*
`