Crypto exchange volume quality
Venue-reported 24h perpetual volume, side by side with what the venue's order books observably support. This dataset does not claim any venue misreports volume; it measures how strongly reported activity is supported by observable signals.
How the signals work
Signals: median spread, slippage at standardized sizes and the largest standardized order that fills on the flagship BTC and ETH markets. A high reported-to-support ratio means reported activity is large relative to visible liquidity — which can reflect genuine high turnover and market-maker flow just as well as anything else. Full method: methodology.
18
Venues assessed
18
Complete data
2026-09-12
Current version observed since
10-25 06:40 UTC
Latest sweep
v3
Methodology · established
Perpetual markets — BTC and ETH
Sources: venue tickers (reported) + PerpFinder order-book engine (observable) · 10 sweeps in the last 3 days| Exchange | Reported 24h (venue) | Reported BTC+ETH | Median spread | Slippage @$100k | Reported / tested fill | Data confidence |
|---|---|---|---|---|---|---|
| Binance | $37.3B | $15.8B | 0.03 bps | 0.00 bps | 15,780×p96 among venues this sweep | 100/100 |
| OKX | $16.5B | — | 0.03 bps | 0.00 bps | — | 80/100 |
| Toobit | $14.2B | $10.3B | 0.03 bps | 0.00 bps | 10,253×p88 among venues this sweep | 100/100 |
| Gate.io | $10.5B | $8.5B | 0.03 bps | 0.00 bps | 8,538×p81 among venues this sweep | 100/100 |
| Bybit | $9.6B | $5.3B | 0.03 bps | 0.11 bps | 5,309×p65 among venues this sweep | 100/100 |
| WhiteBIT | $8.4B | $6.9B | 3.43 bps | 2.60 bps | 6,930×p73 among venues this sweep | 100/100 |
| MEXC | $7.9B | $4.1B | 0.03 bps | 0.28 bps | 4,100×p58 among venues this sweep | 100/100 |
| Bitget | $6.2B | $4.0B | 0.03 bps | 0.00 bps | 4,012×p50 among venues this sweep | 100/100 |
| XT.COM | $5.3B | — | 0.04 bps | 0.00 bps | — | 80/100 |
| BingX | $5.2B | $1.1B | 0.14 bps | 0.11 bps | 1,126×p35 among venues this sweep | 100/100 |
| Bitunix | $3.9B | $2.6B | 0.03 bps | 0.01 bps | 2,590×p42 among venues this sweep | 100/100 |
| KuCoin | $1.9B | — | 0.14 bps | 1.05 bps | — | 80/100 |
| HTX | $1.0B | $868.1M | 0.03 bps | 0.18 bps | 868×p27 among venues this sweep | 100/100 |
| Deribit | $904.2M | — | 0.46 bps | 0.21 bps | — | 80/100 |
| BloFin | $622.3M | $303.9M | 0.06 bps | 0.00 bps | 304×p19 among venues this sweep | 100/100 |
| Kraken | $590.0M | — | 1.13 bps | 0.94 bps | — | 80/100 |
| Phemex | $338.9M | $108.7M | 0.03 bps | 0.84 bps | 109×p4 among venues this sweep | 100/100 |
| Crypto.com | $311.5M | $294.8M | 1.81 bps | 1.30 bps | 295×p12 among venues this sweep | 100/100 |
One value for every venue in this sweep, so it is stated here instead of repeated down a column: max standardized fill ≥ $1.0M · data status complete.
A dash means unavailable. Spread and slippage use two decimal places; 0.00 bps includes readings that round to zero. Data confidence measures our field and market coverage. It does not rate a venue or verify reported volume. A sufficient fill covers at least 99.5% of the tested order within the engine's 2% price band. No adjusted dollar volume estimate is published. Spot is a separate market type and is excluded from this table.
Not assessed
Ranked by reported 24h volume in the PerpFinder daily venue snapshot of 2026-10-25These venues sit in the top 20 by reported 24h perpetual volume, and the perpetual-markets table carries no row for them. A missing row is an omission on our side, never a finding about the venue. A row marked “unverified” is separate: that venue's reported volume is held out of every PerpFinder total, share and ranking, and the reasons are listed with it. Read the rule in the methodology.
| Exchange | Rank by reported volume | Reported 24h (venue) | Why there is no measurement |
|---|---|---|---|
| Backpack | #19 | $96.9M | No measurement for this venue is available in the selected sweep window. |
Reproduce a measured ratio
In the 2026-10-25T06:40:16.207Z sweep, Binance reported $15,779,706,770 of combined BTC and ETH 24-hour volume. Both markets passed the $1,000,000 tested order size.
$15,779,706,770 ÷ $1,000,000 = 15,779.70677 → 15,780× after rounding
The $1 million ladder limit is the largest tested size. A venue that passes it may support larger orders. The ratio uses that tested amount as its denominator; it does not measure the venue's maximum capacity.
The numerator covers 24 hours of reported activity. The denominator comes from sampled buy-side order books. Replenished liquidity and repeated trading can produce a high ratio without any reporting error. Sampled depth does not guarantee a later fill.
Read the latest measured rows as JSON using reportedMainMarketsUsd, maxSufficientFillUsd, and reportedToSupportRatio. The venue volume feed supplies reported activity. Current responses can differ from this saved sweep.
The original dataset began on 2026-07-16. This table uses methodology version 3; earlier versions remain separate. Sweep timestamps describe PerpFinder collection. The public response has no common upstream timestamp for all venue observations.
The same ratio over time
Reported volume vs observable support
6-hour sweepsReported BTC+ETH main-market volume divided by the largest standardized order the venue's visible book absorbed in the same sweep. These are measurements of public data, not accusations — a high ratio does not claim that any venue misreports volume.
Newest stored sweep: 2026-10-25 06:00 UTC, methodology v3. 18 of 18 venues had complete order-book data. 13 venues had a computable ratio. 10 sweeps are stored in the last 3 days.
Top 5 venues by reported 24h volume in the newest sweep that carries a ratio. A sweep with no computable ratio for a venue leaves a break in that line — never a zero.
Why reported volume alone is not enough
Reported volume is self-declared by each venue and is the single most-cited exchange statistic. Order-book depth, spread and slippage are independently observable. Comparing the two does not prove anything about any single venue — but across venues, measured identically and repeatedly, the relationship between reported activity and observable liquidity is itself a useful market statistic. We publish the signals separately rather than collapsing them into one opaque score, and where data is missing we say so instead of guessing.