Every footprint chart in crypto is drawn from one exchange. Binance is the biggest book there is, and it is under a seventh of the BTC perpetual market. XrayCharts merges five books into one ladder, one delta and one CVD, so the size you read is the size that moved price.
No card. Every tool, 30 days of history, delayed 48 hours.
One minute of BTC perpetual. Price wicked down to 63,600 and closed 37 dollars higher. Here is that bar as a single exchange saw it. Press the switch and watch what five books saw.
This is what your chart shows you. The bar closed up, the delta is positive, and the level price bounced off took a few BTC of selling like every other level. There is no reason to look twice at it.
Same minute. Same market. Opposite conclusion.
An illustrative bar rather than a screenshot, drawn to the rules the product draws with, so the levels sum to the volume and the delta printed beside them. Real captures start in the next section.
These are screenshots, not diagrams. Same instrument in each one, same minute resolution, every layer reading the merged tape rather than one exchange's share of it.
Every price, both sides, every bar. Aggressive selling left of the gutter and buying right, each row shaded by its own net delta on a scale shared across everything on screen. Shown with the imbalance outlines and the point of control boxes switched off, because the ladder is the thing worth looking at here. Those are hundreds of BTC a level, because they are the sum of five books rather than one exchange's share of them.
Captured from the running product with the grid and crosshair switched off, and the divergence arrows and unsupported move diamonds drawn at three times their normal size so they read once the picture is scaled to this page. Every number, bar and band on screen is the product's own.
Open is the previous close, and the high and low are the true intrabar extremes of the running total. A bar that ran to +10, reversed to −3 and closed at +2 shows that whole path, which is exactly where absorption lives.
Body size and delta are each divided by their own rolling standard deviation, so one number covers absorbed aggression and moves nothing paid for. Signals never repaint.
Low and high volume prices found by valley prominence rather than by taking the lowest levels, so a small dip inside a heavy shelf is correctly ignored and a genuine seam is not.
Visible range and fixed range, with point of control and a configurable value area. Drag a box and profile that price band only.
Trendlines, rays, zones, Fibonacci and a measure tool that reports change in price, bars, total volume and change in delta. Anchors snap to OHLC and persist per instrument.
Session VWAP with volume weighted deviation bands, plus moving averages, Bollinger, Keltner and Donchian. Any number of instances, each configurable.
One full day of the BTC/USDT perpetual. The number worth looking at is not ours. It is Binance: the largest book in crypto, and still under a seventh of what traded. The only real question is how much of the market your chart gives up.
| Venue | 24h volume | Notional | Share | Of the whole BTC/USDT perp market |
|---|---|---|---|---|
| Binance USDⓈ-M | 147,537 BTC | $9.53B | 14.6% | |
| OKX Swap | 92,341 BTC | $5.96B | 9.2% | |
| Bybit Linear | 60,809 BTC | $3.92B | 6.0% | |
| Gate Futures | 58,486 BTC | $3.78B | 5.8% | |
| Bitget Futures | 52,917 BTC | $3.41B | 5.2% | |
| XrayCharts aggregate | 412,090 BTC | $26.61B | 40.8% | |
| Everywhere else, 40 venues | ≈597,000 BTC | ≈$38.6B | 59.2% |
Those same five books hold 44% of all BTC/USDT perpetual open interest, measured on 1 August 2026. Open interest is capital committed to a position rather than turnover an exchange chooses to report, which makes it the harder number to inflate. By it, the five we read are the deepest in the market.
Our own five are exact: they are that day's executions out of our store, and they match each venue's own published daily turnover to within a rounding error. The market total counts every venue listing a BTC/USDT perpetual whose reported volume is at most three times its own open interest, which excludes about a dozen reporting six to nine hundred times theirs. Two thirds of that total is measured directly from the venues' own APIs; the remainder is venues with no public history, carried at their current share.
A candle tells you where price ended up. A footprint tells you what it cost to get there: how much was lifted at each level, how much was hit, and where the two stopped agreeing.
Market sells hitting the bid at that price, in BTC.
Market buys lifting the offer. Both sides are printed for every price the bar traded at.
The more one side outweighed the other at that price, the stronger the fill. The scale is shared across every bar on screen, so a level looks heavy only if it is heavy compared to what else is in view.
The single price the most volume changed hands at inside that bar.
Buying at one level against selling one tick below it, at three to one or more. The standard test, with a noise floor so a near empty level cannot qualify.
Six price levels of one minute at five dollar buckets, drawn the way the chart draws them. On the aggregate these numbers are the sum of five books, so they are a great deal larger.
Both plans give you the full aggregate across all five venues, the complete history, and every tool on this page. The only thing you pay for is seeing it as it happens.
Everything, 48 hours behind. Enough to study the market properly.
Delayed 48 hoursThe same platform, streaming, with no delay at all.
Real timeStart on Free and upgrade when the delay starts costing you more than the subscription.
On the free plan the chart stops 48 hours behind the present. Everything before that point is the complete aggregate, at full trade level detail, with every tool available. You can study structure, build a read and back check it. What you cannot do is trade the right hand edge.
Because the biggest exchange is about a seventh of the market. Order flow is an argument about who was aggressive and where they were absorbed, and answering it from a seventh of the tape means every threshold you set is calibrated against a sample rather than the population. A level defended across five books and a level defended on one look identical on a single venue chart.
No. Stored executions are never rewritten. The venues genuinely quote a few dollars apart, so the correction is measured from live quotes and applied only when a price ladder is built, onto the same scale the candle is drawn on. Nothing is estimated and nothing is smoothed into place.
The BTC USDT perpetual on Binance USDⓈ-M, OKX swap, Bybit linear, Gate futures and Bitget futures, as one aggregated instrument and as each venue on its own. A venue only joins once it serves an arbitrary historical range over its public API, because a hole that can never be refilled is worse than a venue you left out. More instruments are being added.
Thirty days of trade level order flow, on both plans. Price history reaches considerably further. Where order flow has not been collected the chart marks it, rather than drawing a quiet stretch and letting you believe the market was quiet.
Yes. Every venue is selectable as its own instrument, and the aggregate header shows which books are contributing to the bar you are looking at. Comparing the two is the fastest way to see what your old chart was leaving out.
No. XrayCharts is a data and charting tool. It shows you what traded and where. Nothing on this site is a recommendation, and trading leveraged derivatives carries a real risk of losing more than you put in.
Make a free account and put a real aggregated footprint next to the chart you use now. If the other four books turn out not to matter, it cost you ten minutes and no money.