Over 107 days of 20-level order-book snapshots, xyz:CL quoted a median spread of 0.33 bps in US regular hours with $1.4M resting within 25 bps of mid. A $25k market order paid about 0.74 bps in regular hours and 1.16 bps on weekends, when depth within 25 bps fell to $242k. Taker fee is 0.90 bps with growth mode on; funding averaged -50.19% annualised over the last 30 days.
spread · rth
0.33 bps
median · p90 1.21 bps
depth ±25bps · rth
$1.4M
median, both sides
$25k slip · rth
0.74 bps
fills 100%
$25k slip · wknd
1.16 bps
fills 99%
taker fee
0.90 bps
fee scale 1.0 · growth mode on
funding apr · 30d
-50.19%
longs pay when positive · 7d +1.41%
vol 24h · 30d avg
$231.9M
notional
open interest
$150.7M
both sides · 2026-10-02
by sessionNew York time · medians over 107 days
session
snaps
spread
p90
top
depth ±10
depth ±25
depth ±50
$5k
$25k
$100k
$100k fills
$250k fills
US regular 9:30–16:00
4,902
0.33
1.21
$12k
$1.4M
$1.4M
$1.4M
0.46
0.74
1.24
99%
88%
US extended 4–9:30, 16–20
7,222
0.23
1.09
$9k
$1.2M
$1.2M
$1.2M
0.44
0.75
1.27
95%
75%
Overnight 20–4
4,936
0.24
1.00
$7k
$1.3M
$1.3M
$1.3M
0.44
0.77
1.27
99%
87%
Weekend / holiday
9,154
0.14
0.93
$4k
$242k
$242k
$242k
0.51
1.16
1.67
41%
19%
bps unless stated · RTH 9:30–16:00 · extended 4:00–9:30 and 16:00–20:00 · overnight 20:00–4:00 on weekdays · weekend runs Friday 20:00 to Monday 4:00 and includes NYSE holidays · slip = mean of the buy and sell side of a market order walking the visible 20 levels · fills = share of snapshots the size fits inside those levels · "—" = not fillable in the median snapshot
trade flowtaker side · 24 h to 2026-10-02 00:41 UTC · 30 d = last 30 complete UTC days · all markets
alert melarge orders in CL, $50k or more→ sign infree · by email · when the data is rebuilt
taker buy · 24h
$114.0M
30d $3.45B
taker sell · 24h
$123.4M
30d $3.51B
imbalance · 24h
-4.0%
of $237.4M · 30d -0.8% of $6.96B
trades · 24h
105,365
668 large orders · 30d 2,984,806
takers · 24h
1,866
530 makers · 2,026 addresses
takers · 30d
15,858
16,318 addresses incl. makers
top-10 share · 30d
44%
of $6.96B by 15,858 takers · 15 make half
concentration · 30d
0.035
Herfindahl over takers · 1 = one taker
daily net taker flow
USD per UTC day · taker buy − taker sell · up = net buying, down = net selling · 108 complete days · hover a day for its value · today (2026-10-02) is still open and not drawn: +$759k so far
by session · 30 d
session
taker buy
taker sell
net
imbalance
trades
share of notional
US regular 9:30–16:00
$1.35B
$1.36B
−$7.4M
-0.3%
917,060
39%
US extended 4–9:30, 16–20
$1.12B
$1.14B
−$22.1M
-1.0%
911,106
32%
Overnight 20–4
$497.4M
$521.9M
−$24.4M
-2.4%
482,286
15%
Weekend / holiday
$478.6M
$482.6M
−$4.0M
-0.4%
674,354
14%
sessions in New York time as in the liquidity table above · share = the session's taker notional over all four, $6.96B in 30 days
hour of week · taker notional
New York time · 30 days summed per hour · brighter = more volume · log scale between the 5th and 95th percentile · 168 of 168 hours with trades · hover a cell for USD and trades
large orders: fills grouped by transaction hash, taker and side, $50k or more per order; fills without a hash are not grouped · share = the address's taker notional over all $6.96B of 30-day taker notional (2,984,806 trades, 15,858 takers) · addresses are public on-chain data and link to the Hyperliquid explorer; an address can be a subaccount, vault, market maker or bot and is not labelled · taker notional is volume, not a change in open interest · rebuilt nightly and at 09:05 ET, not a live tape · raw: /api/xyz/flow/CL.json · definitions in methodology
alert meCL funding above or below a level I set→ sign infree · by email · when the data is rebuilt
last settlement
-11.82%
-0.1349 bps/h · 10-01 20:00 ET
next · estimate
-11.04%
-0.126 bps/h · as of 10-01 22:15 ET
apr · 24h
+5.28%
11 of 24 hours observed
apr · 7d
+1.41%
133 of 168 hours observed
apr · 30d
-50.19%
628 of 720 hours · stdev 113%
paid by a long · 30d
-359.82 bps
of notional, over 628 observed hours
hours above neutral · 30d
8%
at 19% · below 72% · negative 67%
range · 30d
+470%
highest hour · lowest -595%
daily funding apr
mean of the observed hourly rates per UTC day, annualised · up = longs paid, down = shorts paid · dashed = neutral 5.48% · 110 complete days · hover a day · today (2026-10-02) is still open and not drawn: -11.82% over 1 h so far
hour of week · mean funding apr
New York time · whole archive · the hour a settlement covers · 168 of 168 hours with data · hover a cell for its mean and n
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Mon
Tue
Wed
Thu
Fri
Sat
Sun
shorts paylongs pay · full shade = ±40.99% APR or more · empty = no settlements
funding by session
session
apr · 30d
hours · 30d
above neutral · 30d
apr · all
hours · all
above neutral · all
US regular 9:30–16:00
-107%
102
0%
-31.42%
416
22%
US extended 4–9:30, 16–20
-69.27%
181
3%
-19.09%
708
22%
Overnight 20–4
-47.17%
131
0%
-12.29%
473
16%
Weekend / holiday
-8.70%
214
21%
-1.14%
862
31%
mean of the hourly rates settled in each session, annualised · 30d = the 720 hours ending at the last settlement, all = since 2026-06-14 · above neutral = share of those hours settled above 5.48% APR
largest settlements · 30 d
settled · ET
session
rate bps/h
APR
09-10 17:0009-10 21:00 UTC
US extended
-6.787
-595%
09-11 17:0009-11 21:00 UTC
US extended
-6.415
-562%
09-10 15:0009-10 19:00 UTC
US regular
-5.818
-510%
09-09 17:0009-09 21:00 UTC
US extended
-5.745
-503%
09-10 14:0009-10 18:00 UTC
US regular
-5.52
-484%
09-14 17:0009-14 21:00 UTC
US extended
-5.429
-476%
09-25 21:0009-26 01:00 UTC
weekend / holiday
+5.365
+470%
09-11 15:0009-11 19:00 UTC
US regular
-5.319
-466%
09-09 16:0009-09 20:00 UTC
US regular
-5.274
-462%
09-10 13:0009-10 17:00 UTC
US regular
-5.227
-458%
positive = longs pay shorts · the neutral rate is the deployer's carry (base interest 0.01% per 8 h × funding multiplier 0.5), not crowd positioning · every mean is over observed hours; collector gaps are not filled · rebuilt nightly and at 09:05 ET, not live · raw: /api/xyz/funding/CL.json · definitions in methodology
slippage curve
median bps over mid vs order size · a line ends where the size stops filling in the median snapshot
hour of week
New York time · median per cell over 107 days · brighter = better · the breathing cell is now
weekly · spread
median bps, all sessions
weekly · depth ±25 bps
median USD both sides, all sessions
cost calculator
per round trip = 2 × median slippage + 2 × taker fee · funding uses the 30-day mean hourly rate, annualised, charged for a position held all day · slippage between tested sizes is interpolated on a log scale · ignores maker rebates, impact that persists after your fill, and fee tiers