Ethereum liquidation heatmap
Where leverage is stacked on ETH perpetuals, the market hit hardest when Bitcoin moves. Bright bands are where forced closes are most likely to fire.
Loading the map Contract Binance USD-M perpetual Updated --
ETH spot
Binance ETH/USDT, last trade
Open interest
Binance USD-M perpetual, notional
24h volume
Binance ETH/USDT, 24 hours
Leverage within 2%
Modelled from open interest
ETH/USDT Liquidation HeatmapBinance USD-M perpetual
The liquidation heatmap for this venue and window, as of the time shown above. Clusters are modelled from open interest and standard leverage tiers, then cleared once price trades through them. Real liquidation prices depend on position size, margin mode and maintenance margin, so read this as pressure, not guaranteed triggers.
Spot -- Short liquidity above -- Long liquidity below --
Ethereum (ETH) Liquidation Levels
How to read this →Above price
Short liquidations --Waiting for the feed.
Below price
Long liquidations --Waiting for the feed.
About this data
Updated Sep 2026Figures are shown as the source publishes them, and anything we calculate is explained below. How the data is made Report an error General information, not financial advice.
What the map shows
A perpetual futures position is borrowed money. Every leveraged trade carries a price at which the exchange closes it automatically, because the margin behind it has run out. That price is not a secret: it follows mechanically from the entry price and the leverage used.
This map is the liquidation heatmap for the selected venue's ETH perpetual over the chosen window. It takes the open interest added in each period, works out roughly where those positions were opened, and projects the price at which each standard leverage tier would be liquidated. Stack all of those projections together and you get a surface: bright where a lot of forced closing sits waiting, dark where there is almost none.
Ethereum is the only asset besides Bitcoin that carries the long windows, and its leverage has a character of its own: deep enough to absorb, but with a higher beta than BTC that means clusters here get reached by moves that started somewhere else.
Time runs left to right, price runs bottom to top: The dotted line is the actual ETH price over the same window. Bands that survive to the right edge are leverage that price has not yet reached.
How to read it
- Bright bands above price are shorts. If ETH rallies into one, those shorts get bought back by the exchange, which adds buying to a move that is already up.
- Bright bands below price are longs. If ETH falls into one, those longs get sold, which adds selling to a move that is already down.
- A band that ends abruptly was consumed. Price traded through it, the positions were closed, and the liquidity is gone. That is why the surface is cleared behind the price line rather than left painted.
- Thin, tight bands very close to spot are the 100x crowd. They are large in count and small in notional, and they get taken out constantly.
- Wide bands far from spot are the 10x crowd. Rarer, much bigger, and the ones that produce the headline cascades.
What is different about Ethereum leverage
ETH perpetuals are the second-deepest leveraged market in crypto, and the only one besides Bitcoin with 180 day and one year windows. The book is thick enough that a small cluster does not move it, but not as thick as BTC, so the same notional of forced flow travels further in price here than it would on Bitcoin.
The defining feature of Ethereum leverage is that it is often liquidated by a move it did not start. ETH trades with a higher beta than BTC in both directions, so a cascade that begins on the Bitcoin book arrives on the Ethereum book amplified. A cluster on this map that looks comfortably far from spot can be reached by a Bitcoin move of half the size. Reading this page without the BTC map open beside it misses where the risk usually comes from.
Not all ETH leverage is a directional bet on Ethereum. A meaningful share is the ETH/BTC ratio trade, where one leg is long ETH and the other short BTC, and carry positions built around staking yield. Those show up on this surface as ordinary long or short leverage, but the trader behind them is exposed to a spread, not a price, and will manage them differently when the level is approached.
Since the spot ETF launch, Ethereum has an unleveraged flow that runs alongside the derivatives book. Sustained ETF inflows put spot bids under price that a cascade has to trade through, and outflows remove them. The ETH ETF dashboard on this site is the other half of the picture when a large cluster below spot is in play.
Use the long windows on ETH
Ethereum carries the 180 day and one year maps on Binance, Bybit and OKX. The wide bands that survive on those windows are low-leverage positions that have sat through months of price action, and they are where Ethereum's largest single-day liquidation events have been triggered. They are invisible on the 24 hour view.
Why clusters move price
A liquidation is not a normal sale. It is a market order the exchange sends on the trader's behalf, at whatever price the book offers, with no regard for slippage. When a dense cluster is hit, thousands of those orders fire inside a few seconds against a book that has usually already thinned out.
That is the cascade mechanic. Price reaches a cluster, the cluster fires, the firing pushes price further in the same direction, and the push reaches the next cluster. It stops when it runs out of stacked leverage or when resting bids and offers are deep enough to absorb the flow.
Why clusters attract price
This is also why clusters act like magnets in quiet markets. A pool of guaranteed market orders sitting at a known price is a target, and desks that can see it have every reason to push toward it.
Using it without getting run over
- Keep the Bitcoin map open: Ethereum's clusters are usually reached by a BTC move, amplified. Judge the distance to an ETH level in Bitcoin terms as well as in ETH terms.
- Treat clusters as risk, not as entries: Knowing where forced selling sits tells you where a move is likely to accelerate. It does not tell you the move is coming.
- Do not park a stop inside a cluster: That is precisely the price where slippage is worst.
- Watch which side is heavier: When one side of price holds most of the leverage, the path of least resistance usually runs toward it.
- Check the window: A cluster built over thirty days is a different animal from one built in the last six hours.
- Size for the cascade, not the level: If you are long into dense long liquidity below, assume the fall through it will be faster than normal.
What it cannot tell you
- A blended level is not one order book: All exchanges adds the venues together, so a bright band there can be mostly one book. Only that book's own price trading through it triggers those positions, so pick the venue out before sizing around a level.
- It cannot see cross-margin: A trader with collateral spread across several positions is liquidated on portfolio health, not on a single price.
- It assumes standard leverage tiers: Real traders use 7x and 33x and every number in between.
- It is not a forecast: A dense cluster at 8% below spot says what happens if price gets there. It says nothing about whether it will.
Not investment advice
This page describes market structure, not what to do about it. Nothing here is a recommendation to buy or sell anything. Leveraged trading can lose you more than you put in, and a map of where other people are positioned is not a reason to take a position yourself.
Questions
How often does this heatmap update?
Every 60 seconds while the page is open. The status line under the headline shows when the map was last updated. The figures beside it refresh on the same cadence, and each one says underneath where it came from.
All exchanges, or one at a time?
They answer different questions, so the page opens on the blend and keeps the books beside it. All exchanges: where the leverage sits market-wide, which is what you want when you are asking whether a level matters at all. One venue: whose book it sits in, which is what you want before trading on it, because a cluster on one exchange is only reached when that exchange's own price gets there. A band that looks large on the blend can be one venue carrying almost all of it.
Are these real liquidation orders?
No, and no public heatmap is. Exchanges do not publish the liquidation price of every open position. What is real here is the price data and the open interest that feeds the model. The projection from open interest onto liquidation prices is a model, and what it cannot show is set out under what it cannot tell you.
Can it predict where Ethereum goes next?
It cannot. It maps where an ETH move would accelerate, not whether one is coming, and on Ethereum the move that gets it there very often starts on the Bitcoin book. Its value is in sizing and in stop placement, with the BTC map open beside it.
What does clearing a cluster mean?
Once price trades through a level, the positions that would have been liquidated there are gone. Leaving the band painted would show liquidity that no longer exists, so the map wipes any cluster the price has crossed. That is why the surface behind the price line is dark.
Which window should I use?
The 24 hour view for intraday work, where the clusters that matter were built in the last few sessions. The 7 and 30 day views for position sizing. The 90 day, 180 day and one year views for the large low-leverage bands that have survived months without being touched.
Ethereum has all seven windows on Binance, Bybit and OKX because the 180 day and one year maps are published for ETH and BTC. Hyperliquid stops at 90 days. The long windows are where Ethereum's largest, slowest-built clusters live.
Why do ETH clusters fire when Bitcoin moves?
Ethereum trades with a higher beta than Bitcoin, so a move on the BTC book arrives on the ETH book amplified. A cluster that sits 5% from ETH spot can be reached by a Bitcoin move of roughly half that size. That is why the risk to Ethereum leverage usually originates on a different map.