PerpRadar
as of 00:49 UTC

Find perp setups that actually pay — 19 venues, measured live.

Setups ranked by what you’d actually keep: net APY after estimated fees and slippage, sized to the notional you pick — across venues most aggregators don’t track. Annualized rates only on proven funding intervals; gaps stay labeled gaps.

$13.3Btotal OIi
19venuesi
791assetsi
1961marketsi
$13.1Btracked 24h volumei
Since yesterday2026-09-10 00:49 UTC

Funding movers · 24h

Largest recorded 24h changes in annualized funding — was → now from PMR’s own history, proven intervals only, across the 45 markets on today’s radar.
  • LAPTOP @ aster
    flipped 7h ago Funding is shorts pay longs now
    +11.0%-11012%
  • H100 @ lighter
    flipped 40m ago Funding is shorts pay longs now
    +429%-927%
  • AI @ lighter
    Funding is longs pay shorts now
    +1507%+200%
  • BMT @ aster
    flipped 6h ago Funding is shorts pay longs now
    +11.0%-530%

Every rate is annualized only on a proven funding interval; the sign is who pays whom, never a call.

Predictions watch

template:sportsContestWinner
≈24h ago 60.9%+38.4 ptsnow 99.2%
Latest settled: template:binaryPrice — last recorded 2.0% · 152 resolved in 24h

Probabilities are the venue’s own recorded mids on books PMR records.

Setup of the day

ANSEM 2 venues carry it
Annualized funding lighter +253% vs aster +15.9%
237% net carry · annualized
Fill cost · both legs$5k≈$116.95233.9 bps$10k≈$333.87333.9 bps$25k
Swings a lot
0 flips/24h · APY -183–671%

Picked from 12 ranked setups — the same builder the Funding workbench uses; measured books and a steady day lead.

Watchlist and alerts

Track the markets you care about and get a heads-up when a setup appears — e.g. “SOL funding above 50% annualized on a trusted venue.”

Alerts deliver via Telegram once connected — suppressed or stale data never fires.

Venue scoreboard

The venue reference ranked by 24h volume — with the measured cost of a $25k fill on books PMR captures, open interest and baseline fees alongside.

1 Hyperliquid 13.3 bps $6.79B $10.7B 0.015% 0.045%
2 Aster 66.0 bps $1.78B $1.09B 0.01% 0.035%
3 Lighter 23.3 bps $1.3B $460M 0% 0%
4 ApeX Omni 16.8 bps $974M $135M 0.02% 0.05%
5 Pacifica 14.6 bps $816M $90.7M 0.015% 0.04%
6 Extended 57.3 bps $498M $156M 0% 0.025%
7 GRVT 12.4 bps $496M $410M −0.0001% 0.045%
8 Backpack 7.9 bps $284M $93.7M 0.02% 0.05%
9 Orderly $76.2M $33.2M 0% 0.03%
10 dYdX 131.8 bps $22.1M $36.2M 0.02% 0.05%
11 Derive $14.2M $55.9M 0.01% 0.03%
12 Hibachi 24.1 bps $7.98M $1.5M 0% 0.045%
13 Aevo 96.1 bps $7.89M $11.3M 0.05% 0.08%
14 Paradex 319.2 bps $5.21M $9.3M 0% 0.0075%
15 Vest 26.1 bps $5.18M 0.01% 0.01%
16 Bluefin 13.3 bps $1.34M $843K 0.01% 0.035%
17 edgeX $0 $1.57M 0.015% 0.038%
18 Ethereal $0 0% 0.03%
Helix 54.9 bps 0.0001% 0.0001%

Ranked by reported 24h volume, summed across each venue's in-scope markets — a venue reporting no volume shows “—” below the ranked rows: a gap, never a penalty. Any column header re-sorts the table (a venue missing that figure sorts last; “#” keeps the volume rank). $25k fill · median is the measured cost of crossing the venue's captured books at that size (the worse side of each walk); “—” means outside the current book-capture budget. Maker / taker are baseline retail fees (as of 2026-06) — rough context, not exact cost; hover a fee for its source. A labeled “n/a” is an honest gap, never invented.

Proven intervals, or no APY

An annualized rate appears only when a venue’s funding cadence is proven from its own data. Unproven interval → the APY is suppressed, visibly — never guessed.

Costs at your size

Setup economics are sized to the notional you pick on the Funding workbench. Fees and slippage are labeled estimates today — the book depth our collectors already capture is being wired into them.

Gaps stay gaps

A number a venue doesn’t report stays “—”. Nothing is zero-filled, interpolated, or scraped from a third party.

Recording what others didn’t

Funding, open interest and book depth on these venues, captured by our own collectors — months before most started tracking them. History like that can’t be backfilled by anyone who wasn’t recording.

Learn: guides to reading perp-DEX data · Questions, ideas, something broken? Message us on Telegram — it lands with the operator.