Hyperliquid runs a fully on-chain order book for perpetual futures, and its base taker fee of 0.045% is a little under Binance's 0.05%. That makes it look cheaper on paper. But fees are one of three costs. We put Hyperliquid and Binance side by side on the same market orders, on the depth of their books, and on almost three and a half years of funding payments.
Fees: Hyperliquid is a little cheaper at every tier we checked
| Level | Hyperliquid | Requirement | Binance | Requirement |
|---|---|---|---|---|
| Base / VIP 0 | 0.015% / 0.045% | base | 0.02% / 0.05% | regular account |
| Tier 1 / VIP 1 | 0.012% / 0.04% | 14-day weighted volume over $5M | 0.018% / 0.05% | 30-day futures volume ≥ $5M and ≥ 5 BNB |
| Tier 2 / VIP 2 | 0.008% / 0.035% | 14-day weighted volume over $25M | 0.016% / 0.04% | 30-day futures volume ≥ $10M and ≥ 25 BNB |
| Tier 3 / VIP 3 | 0.004% / 0.03% | 14-day weighted volume over $100M | 0.012% / 0.032% | 30-day futures volume ≥ $50M and ≥ 100 BNB |
On a $100,000 market order, the base fee is $45.00 on Hyperliquid and $50.00 on Binance. The tiers are not the same ladder: Hyperliquid's first step needs $5M of 14-day volume, while Binance's VIP 2 (the first level that lowers the taker fee) needs $10M of 30-day futures volume and 25 BNB. Hyperliquid also offers staking discounts and a referral discount on fees; Binance offers a BNB discount. None of those are in the numbers here.
Order books: Binance was cheaper to cross
Base-tier taker fee plus median hourly spread and slippage, basis points, 6 Nov 2026, 00:00 UTC to 10 Nov 2026, 03:00 UTC.
Show the numbers
| Size | Hyperliquid spread+slip | Binance spread+slip | Hyperliquid total | Binance total |
|---|---|---|---|---|
| $1k | 0.08 bps | 0.01 bps | 4.6 bps | 5.0 bps |
| $10k | 0.10 bps | 0.01 bps | 4.6 bps | 5.0 bps |
| $100k | 0.18 bps | 0.02 bps | 4.7 bps | 5.0 bps |
| $500k | 0.39 bps | 0.17 bps | 4.9 bps | 5.2 bps |
| $1M | 0.61 bps | 0.42 bps | 5.1 bps | 5.4 bps |
Binance's BTC book is tighter at the top and deeper behind it. A $100,000 BTC market order paid a median of 0.02 bps in spread and slippage on Binance and 0.18 bps on Hyperliquid. That gap, about $1.52, is smaller than the $5 fee gap, so Hyperliquid's total was lower in 98 of 100 hours. At $1,000,000 the book gap grows to $18.60 and the totals end up 5.1 bps against 5.4 bps.
| Coin | Hyperliquid | Binance | Hours cheaper (HL / Binance) |
|---|---|---|---|
| BTC | $46.76 | $50.24 | 98 / 2 |
| ETH | $48.87 | $50.84 | 94 / 6 |
| SOL | $52.25 | $57.80 | 83 / 0 |
| XRP | $61.75 | $62.36 | 42 / 41 |
ETH looked like BTC. SOL is the odd one: Hyperliquid's SOL book was tighter at small sizes (0.11 bps against 0.56 bps at $1k), while Binance's was deeper at $500k and above. XRP was close all the way and split the hours almost evenly.
In our live snapshots (13 reads on 10 Nov 2026), Hyperliquid's BTC book held about $43.7M within 0.1% of the mid price on both sides together. For Binance futures we could only see about the first $2.5M per side, all of it within 0.019% of the mid, so its depth within 0.1% was at least $5.1M. Hyperliquid's figure includes the grouped (coarse) levels described in the method, which flatter it slightly. Our liquidity ranking compares depth across all exchanges.
Funding: where Hyperliquid cost much more
Perpetual futures have no expiry, so they use funding to stay close to the spot price: when the perp trades above spot, longs pay shorts, and the other way round. Binance settles every 8 hours; Hyperliquid settles every hour, at one-eighth of an 8-hour rate. The size of the payment is what matters, and over the long run Hyperliquid's was clearly bigger.
Average funding rate each month, turned into a yearly percentage (positive = longs pay). Hyperliquid hourly and Binance 8-hourly data, 12 May 2023 to 31 Oct 2026.
Show the numbers
| Month | Hyperliquid | Binance |
|---|---|---|
| 2023-05 | -98.7% | 6.5% |
| 2023-06 | 20.5% | 7.3% |
| 2023-07 | 19.9% | 6.6% |
| 2023-08 | -8.3% | 6.3% |
| 2023-09 | -0.3% | 1.2% |
| 2023-10 | 18.3% | 4.4% |
| 2023-11 | 25.5% | 12.4% |
| 2023-12 | 40% | 19.9% |
| 2024-01 | 28.6% | 13.2% |
| 2024-02 | 30.1% | 16.5% |
| 2024-03 | 66.8% | 37% |
| 2024-04 | 18.8% | 10.2% |
| 2024-05 | 16% | 7.5% |
| 2024-06 | 19.7% | 10.3% |
| 2024-07 | 12.3% | 7% |
| 2024-08 | 11.7% | 1.4% |
| 2024-09 | 12.9% | 1.9% |
| 2024-10 | 12% | 8% |
| 2024-11 | 32.6% | 15.8% |
| 2024-12 | 28% | 14.2% |
| 2025-01 | 16.6% | 8.5% |
| 2025-02 | 10.9% | 5.1% |
| 2025-03 | 3.1% | 2.1% |
| 2025-04 | 5.3% | 2.6% |
| 2025-05 | 16.2% | 5.1% |
| 2025-06 | 8.5% | 2.9% |
| 2025-07 | 19.4% | 8.6% |
| 2025-08 | 11.4% | 7.1% |
| 2025-09 | 10.1% | 5.3% |
| 2025-10 | 10.7% | 3.6% |
| 2025-11 | 9% | 5.4% |
| 2025-12 | 6.2% | 5% |
| 2026-01 | 7.3% | 5.4% |
| 2026-02 | 2.3% | -0.8% |
| 2026-03 | 0.9% | -1.1% |
| 2026-04 | -0.8% | -2.2% |
| 2026-05 | 3.1% | 2.7% |
| 2026-06 | 5.9% | 2.5% |
| 2026-07 | 9.5% | 6.7% |
| 2026-08 | 8.3% | 7.3% |
| 2026-09 | 8.8% | 5.5% |
| 2026-10 | 8.8% | 4.8% |
| Year | BTC Hyperliquid | BTC Binance | ETH Hyperliquid | ETH Binance |
|---|---|---|---|---|
| 2023 | 8.7% | 5.2% | 12.9% | 5.3% |
| 2024 | 24.2% | 12% | 22.1% | 13% |
| 2025 | 10.6% | 5.1% | 8.5% | 4.9% |
| 2026 | 4.5% | 2.6% | 4.9% | 1.7% |
Over the whole window (12 May 2023 to 31 Oct 2026), the funding rates add up to 48% on Hyperliquid BTC and 24.9% on Binance BTC; for ETH, 48.4% against 25%. In plain money: a $100,000 BTC long kept open the whole time would have paid about $48,050 in funding on Hyperliquid and $24,910 on Binance (a short would have received those amounts). That gap is hundreds of times larger than the $5 fee difference on one trade.
Both formulas have a small fixed interest part (0.01% per 8 hours, which Hyperliquid pays out hourly); the difference comes from the premium part, which follows how far each perp trades above or below its index. On Hyperliquid that premium ran higher in most months. One likely reason is that its traders lean long more often, with fewer large players on the other side, but the data only shows the result, not the cause. It was not true every month: Hyperliquid's monthly rate was lower than Binance's in 3 of 42 months. Over the period as a whole, a long paid more on Hyperliquid, and a short was paid more.
Which one is cheaper for you?
For a trader who opens and closes BTC or ETH positions within the day, at up to a few hundred thousand dollars, Hyperliquid came out a little cheaper: its fee saving was larger than its wider book. For orders near $1M, Binance's deeper book wins. For anyone who holds a long for weeks, funding dominates everything else, and over the last three years it was roughly 1.9 times as expensive on Hyperliquid. Shorts were on the other side of that trade.
Two things are not in these numbers. Hyperliquid is a decentralised exchange, so deposits and withdrawals go through a bridge and an on-chain wallet, with their own small costs and risks. Binance is a centralised exchange with its own account and custody terms. Neither is advice; they are different products with different risks. You can run today's comparison for your size on the Binance vs Hyperliquid page.
Method
Order-book cost: CostDepth samples both perpetual books every 5 minutes and stores hourly averages of the spread and slippage of market orders from $1k to $1M (6 Nov 2026, 00:00 UTC to 10 Nov 2026, 03:00 UTC). We used the median hour, buy and sell averaged. Hyperliquid's public book gives 20 price levels per side; CostDepth adds a second, coarser read (prices rounded to 4 significant figures) for depth further out, and drops the first coarse bucket to avoid counting liquidity twice. Grouped levels can sit up to one bucket closer to the mid than the real orders, so Hyperliquid's large-order figures are slightly on the kind side.
Depth snapshots: 13 reads of both books between 10 Nov 2026, 03:39 UTC and 10 Nov 2026, 04:34 UTC. Hyperliquid was read directly. Binance futures refuse our research server's location, so its book came from CostDepth's own live book endpoint (fetched from Frankfurt), which keeps about $2.5M per side: Binance depth figures are a lower limit, marked "at least".
Funding: Binance BTCUSDT and ETHUSDT funding from Binance's public data archive (data.binance.vision), one row per 8-hour period. Hyperliquid BTC and ETH funding from its public fundingHistory API, one row per hour, starting 12 May 2023 (its first record). Both were cut to the same window, 12 May 2023 to 31 Oct 2026. The total is the plain sum of the rates: what a long position of fixed size paid, as a share of that size, if held the whole time (a short received it).
Fees are base-tier rates from CostDepth's fee files (Hyperliquid docs and binance.com fee page, 5–6 November 2026). Hyperliquid's own tiers depend on 14-day volume and can be lowered by staking HYPE; Binance's on 30-day volume plus BNB held.
The numbers behind every chart and table are in data.json (free to reuse with a credit to CostDepth). This is fee and cost data, not financial advice.