Real liquidation prices of tracked Hyperliquid positions, in USD. Longs liquidate below the mark, shorts above.
ETF perp on Hyperliquid (deployed by xyz)
Mark 81.646Mapped $134K
Bright bands = where positions get liquidated. Every cell is the USD of real Hyperliquid positions whose liquidation price (as Hyperliquid reports it) sits in that price range at that time. Brighter means more. Green = longs (below the price), red = shorts (above it).
Candles = price. The column on the right is the liquidation map of the newest snapshot: bars per price level, lines = cumulative USD liquidated if the price moves there.
Price moving into a bright band can force-close those positions at once. Hatched columns = no snapshot stored.
Click a cell to list the tracked wallets with their liquidation price in that band now.
Open interest added at this price (optional layer, off by default): where open interest grew on Binance + Bybit while the price traded in each band, from the open interest and price the exchanges publish, in 5-minute steps. It shows where contracts were opened, not their leverage or liquidation price, and not who bought: every new contract has a long and a short side.
Drag to move in time, shift + drag to zoom into a range, Ctrl or ⌘ + scroll to zoom; drag the price axis to stretch it; double-click to reset. Click a cell for the wallets with their liquidation price there.
Free: the last 24 hours of every coin. Up to 1 year of history and 5m to 1D columns are part of Pro.Pro opens soon: 1 year of heatmap history
Latest snapshot 2026-11-05 13:15 UTC, built from 6 tracked TLT positions with a liquidation price within ±50% of the price (snapshots every 5 minutes; top 20 coins also every minute in 0.1% buckets). Wallets are polled in rotation, so positions are minutes old.
| bucket | vs mark | longs | shorts |
|---|---|---|---|
| 93.000–93.500 | +14.2% | - | $57.3K |
| 102.50–103.00 | +25.8% | - | $49.1K |
| 106.00–106.50 | +30.1% | - | $9.4K |
| 67.000–67.500 | −17.6% | $5.3K | - |
| time (UTC) | bucket | longs | shorts | mark then |
|---|---|---|---|---|
| 11-04 14:00 | 89.000 | - | $2.8M | 81.135 |
| 11-04 15:00 | 89.000 | - | $2.8M | 81.135 |
| 11-04 16:00 | 89.000 | - | $2.8M | 81.135 |
| 11-04 17:00 | 89.000 | - | $2.8M | 81.135 |
| 11-04 18:00 | 89.000 | - | $2.8M | 81.135 |
| 11-04 19:00 | 89.000 | - | $2.8M | 81.135 |
| 11-04 20:00 | 89.000 | - | $2.8M | 81.135 |
| 11-04 21:00 | 89.000 | - | $2.8M | 81.135 |
| 11-04 22:00 | 89.000 | - | $2.8M | 81.135 |
| 11-04 14:00 | 76.000 | $1.8M | - | 81.135 |
| 11-04 15:00 | 76.000 | $1.8M | - | 81.135 |
| 11-04 16:00 | 76.000 | $1.8M | - | 81.135 |
Positions near liquidation: TLT whales →
Every tracked position reports its own liquidation price; we add up their USD per price bucket and store that ladder every 5 minutes (the top 20 coins every minute), so each column is a measured snapshot, never an estimate. Columns before our live collection (2026-11-03 14:15 UTC) or in a gap show the daily snapshot of every open Hyperliquid position. Full method · How old the positions are
Free: every coin, the last 24 hours with 15m · 1h columns. Pro: 5m · 4h · 1D columns and up to 1 year. Desk: 1m columns (top 20 coins) and up to 3 years. The 1D heatmap reaches back to 18 Sept 2026 (a daily snapshot of all TLT positions); the intraday history of tracked positions is stored since 23 Oct 2026 and grows every day. Not investment advice.
From the open positions of tracked Hyperliquid wallets, each with the liquidation price Hyperliquid reports for it. The Lighter and Lighter RH views add tracked accounts with the liquidation price the exchange reports. The dYdX and GMX views add every open position read from the chain, with a liquidation price we compute from the on-chain state. No leverage is assumed.
The current map is rebuilt every few minutes; the heatmap over time stores one snapshot every 5 minutes (top 20 coins: every minute, finer price buckets). Large wallets are polled about every 2.5 minutes.
Centralised exchanges do not publish positions, so their liquidation levels are not shown. We do not model them.