Perpetual funding rates by exchange

Median funding, coverage and the spreads each venue takes part in
Exchanges34with live funding
Venue · market rows4,920compared
Exchange pairs212with live spreads

On 3 of 29 fully covered venues the median is exactly zero — Ostium, Risex, dYdX — because more than half of their markets pay no funding at all right now. Among the venues that do charge, the lowest median is Paragon at -103.2%, then trade.xyz (-7.8%) and Nado (-4.9%). The dearest is ApeX at +10.9%. 3 of 29 fully covered venues currently sit below zero, meaning shorts pay longs on most of their markets. A low median makes a venue the cheaper place to be long; the opposite end is where a delta-neutral position wants its short leg.

All exchanges

by markets covered
ExchangeMarketsMedian APRPaying positiveVolume 24hOpen interestBest pair from here
BybitCEX8h · 3 intervals · 30 pairs444+10.9%263 / 444$12.5B$12.3BMEW · +474.8%
BitgetCEX8h · 3 intervals · 30 pairs442+10.9%262 / 442$8.3B$8.2BMEW · +469.1%
Gate.ioCEX8h · 3 intervals · 30 pairs423+1.5%221 / 423$14.7B$13.2BMEW · +474.8%
KuCoinCEX4h · 3 intervals · 30 pairs409+10.9%255 / 409$2.0B$2.8BCASHCAT · +247.9%
MEXCCEX4h · 3 intervals · 30 pairs399+7.1%255 / 399$11.3B$12.3BUNITREE · +439.2%
OKXCEX8h · 3 intervals · 30 pairs331+0.3%168 / 331$22.5B$6.8BMEW · +469.1%
BingXCEX4h · 3 intervals · 28 pairs331+10.9%280 / 331$6.6B—MEW · +435.0%
BinanceCEX4h · 3 intervals · 27 pairs324+8.4%222 / 324$34.3B—MEW · +461.8%
AsterDEX4h · 3 intervals · 30 pairs239+10.9%153 / 239$950.7M$1.1BDOOD · +265.3%
HyperliquidDEX1h · 29 pairs225+10.9%156 / 225$8.7B$14.7BUNITREE · +438.9%
VariationalDEX8h · 3 intervals · 31 pairs210+9.3%127 / 210$3.2B$502.1MGOOGL · +170.8%
ExtendedDEX1h · 30 pairs150+5.3%98 / 150$568.3M$190.1MUNITREE · +438.6%
LighterDEX8h · 30 pairs135+10.5%108 / 135$2.2B$557.0MPRL · +321.8%
HTXCEX8h · 3 intervals · 30 pairs129+10.9%82 / 129$1.2B$2.1BBTW · +196.4%
trade.xyzDEX1h · 23 pairs89-7.8%38 / 89$3.1B$3.2BUNITREE · +436.7%
NadoDEX24h · 29 pairs72-4.9%32 / 72$376.4M$80.0MCASHCAT · +249.8%
Ondo PerpsDEX1h · 29 pairs66+5.5%62 / 66$381.3M$108.9MGOOGL · +164.8%
BackPackDEX1h · 27 pairs64+10.9%51 / 64$268.0M$114.5MZEC · +41.3%
ArcusDEX1h · 28 pairs61+4.4%60 / 61$381.7M$110.4MGOOGL · +167.1%
VestDEX1h · 29 pairs56+5.0%50 / 56$99.7M$45.8MARM · +72.9%
PacificaDEX1h · 29 pairs53+10.9%48 / 53$844.1M$121.5MDRAM · +220.3%
GMTradeDEX1h · 28 pairs49+0.1%30 / 49$126.9B$20.0BGOOGL · +171.5%
OstiumDEX1h · 28 pairs48+0.0%15 / 48—$6.3MGOOGL · +157.3%
RisexDEX1h · 29 pairs43+0.0%18 / 43$117.4M$80.4MGOOGL · +171.4%
SodexDEX1h · 29 pairs36+5.5%30 / 36$221.8M$314.8MCL · +45.9%
dYdXDEX1h · 27 pairs24+0.0%10 / 24$73.0M$49.6METH · +21.9%
ParagonDEX1h · 16 pairs18-103.2%7 / 18$5.4M$14.7MTER · +413.0%
ApeXDEX1h · 27 pairs14+10.9%14 / 14$308.5M$143.7METH · +12.9%
ParadexDEX8h · 26 pairs13+6.2%10 / 13$7.8M$11.9MBTC · +14.0%
EntropyIODEX1h · 17 pairs9+5.5%6 / 9$100.0M$77.2MPRL · +321.8%
01DEX1h · 27 pairs5+21.9%5 / 5$2.2M$1.1MBTC · +12.8%
HibachiDEX8h · 1 pairs4+7.5%3 / 4$31.9M$1.9MHYPE-USDT · +9.2%
GRVTDEX4h · 2 intervals · 1 pairs4+10.9%3 / 4$471.3M$445.5MHYPE-USDT · +9.2%
Markets by KinetiqDEX1h · 1 pairs1+28.0%1 / 1$4.0M$7.9MUS500 · +26.3%

Median rather than average funding: one stale or mis-scaled market is enough to drag an average into nonsense, while the median describes what the typical market on that venue actually pays. Venues with fewer than 10 markets or fewer than 5 counterpart exchanges still have a page, but it is left out of the index — there is not enough there to be worth a search result.

Why funding differs between exchanges

Funding is set per venue, from that venue's own perp-versus-spot basis and its own formula: different intervals (1h, 4h, 8h), different clamps and different premium windows. The same coin can pay positive funding on one exchange and negative on another at the same moment — which is the entire basis of a delta-neutral funding trade.

Rates are annualized before comparison

Venues settle funding on different intervals — from one hour to eight — and the raw per-interval number is not comparable across them. See the exchange-pair pages for the spreads those differences produce, or the methodology for how the numbers are computed.