Market Intelligence

Liquidations

Realized-liquidation data (long/short USD over 1h/4h/12h/24h windows). Defaults to Hyperliquid perps coins only.

Coverage: totals aggregate our configured realized-liquidation venue feeds (OKX + Bybit + Hyperliquid where armed; the HL contribution is exact per-fill capture) — see the coverage field on the response. Binance, the largest venue, geo-blocks its liquidation stream from our infrastructure (no REST fallback), so these figures run below a true all-exchange total.

Filter semantics: exchange= is a coin-set filter, not an exchange-data filter. With exchange=hyperliquid you get the coins listed on Hyperliquid, but the liquidation values inside each entry are the venue-subset totals above. For Hyperliquid's own per-asset trading data use /daily/hyperliquid.

AsterDEX coverage: With exchange=asterdex the response is the union of (a) CoinGlass cross-exchange aggregate rows whose symbol is also listed on AsterDEX and (b) AsterDEX venue-native rows (source: "asterdex") for AsterDEX-only symbols (e.g. LAB, ASTEROID). Native rows depend on AsterDEX's force-order feed, which is intermittently disabled upstream; when unavailable, only the CoinGlass-aggregate rows are returned.

as_of (unix seconds) is when the liquidation snapshot behind data was last refreshed. Each row also carries liq_1h_vs_7d_median {long, short, total, hours_7d}: the trailing hour's liquidated notional over the median hourly notional of the prior 7 days for that coin (1.0 = a normal hour), computed from the exact Hyperliquid fill tape on both sides of the ratio — see baseline for the exact definition. Null per side when the 7d median is zero, per row when fewer than 24 hours of tape exist, and for every row on a box without the backtesting archive.

Row grain: each row's USD fields are trailing-window LEVELS (see grain: kind: rolling_window, sum_ok: false) — polling this every few minutes and summing or differencing the answers double-counts. For incremental flow use the Pro /backtesting/hl-liquidation-bars, built from the Hyperliquid fill tape. count is the number of liquidation events in the 24h window and source the venues they came from (row_count_meaning).

Order: rows are sorted by liquidation_usd_24h, largest first, so limit keeps the coins with the most liquidated notional. AsterDEX-native rows (exchange=asterdex) follow the aggregate rows.

The free tier is scoped to BTC; full-universe coverage is Pro and Pro Plus.

get/api/v1/market-intelligence/liquidations

Request

  • Base URL: https://cryptodataapi.com
  • URL: https://cryptodataapi.com/api/v1/market-intelligence/liquidations
  • Auth: API key in header X-API-Key

Query parameters

symbolstring

Filter to a single coin (e.g. BTC)

exchangestring

Filter to coins on exchange: hyperliquid, binance_spot, asterdex, or all

typestring

Instrument type — currently a no-op reserved for future filtering; only perps data is returned

limitinteger

Max coins to return (default 250)

Response

Successful Response

dataunknown required
as_ofinteger nullable

Unix seconds the liquidation snapshot was refreshed

coverageobject nullable
baselineobject nullable

Definition of the per-row liq_1h_vs_7d_median ratios

grainobject nullable

What a row's USD fields ARE: rolling trailing-window levels (kind=rolling_window, sum_ok=false) — never sum or difference successive polls

row_count_meaningstring nullable

What each row's count counts (liquidation events in the 24h window)

Changes

No recorded changes to this operation across all 1 revision of this API.