Data Accuracy
Accurate and fast are easy to claim. This page shows how we check both: every hour we re-fetch a sample of markets from each exchange's own public API and compare it with what this site shows, every night we re-check yesterday's stored history, and the engine measures how long each feed takes to arrive.
Over the last 7 days, 97.4% of 21,077 scored checks of open interest, funding and price were within tolerance of the exchange's own figure.
How each number is made
Sources. Every figure comes from each exchange's own public market-data API: nothing from paid vendors or other data sites. Prices and volume come from the exchanges' 1-minute bars or tickers, open interest and funding from their REST or WebSocket feeds.
Units. Open interest is in US dollars at the current mark price, and is one-sided on every exchange: the value of all longs, which equals all shorts. Where an exchange reports both sides together (Hyperliquid, Bitget, Gate, Bybit inverse) we halve it. Funding is shown normalised to 8 hours; the checks convert it back to each exchange's own interval, so a wrong interval shows up as a gap.
Combined open interest. MEXC's reported open interest moves about 6.8 times more than its own trades imply, so it is shown per market but not added to combined totals. Those markets still show their own open interest, marked *, in each coin's exchange table. Their open interest is also left out of the weights of each coin's OI-weighted funding average; their funding rate still shows per market and counts in the funding spread.
Liquidations. Binance liquidations are sampled: Binance publishes at most one per symbol per second, so its totals are a floor. Hyperliquid has no public liquidation stream. MEXC and KuCoin publish no liquidation feed at all. Which exchanges are complete is listed on data coverage.
How we check. Every hour a job reads what this site serves and re-fetches the same markets from each exchange's API: a sample of about 60 markets across all venues, always including every market of the five largest coins, plus the open interest totals of the ten largest coins. Every night a second job re-derives yesterday's stored open interest, funding settlements, 1-minute bars and liquidation totals from the exchanges' own history endpoints.
Tolerances. A check is within tolerance when the gap is at most:
- Open interest: 1.00% relative (Binance 2.00%, KuCoin 1.50%, MEXC 1.50%: slower OI polling)
- Funding rate: 0.50 bp of the exchange's native-interval rate
- Price: 0.20% relative, vs the exchange's mark or last price
- Coin totals: 2.00% relative, our coin OI vs the sum of the exchanges' own values
Not scored.
- Not scored, under review: MEXC's reported open interest moves about 6.8 times more than its own trades imply. Until we know what the figure measures, matching it would only show we copy MEXC's number, not that the number is right.
- Not scored: KuCoin's public open-interest endpoints return values from inconsistent caches, often 10-100% apart within seconds, so a gap from them says little about our figure.
- Aster and Lighter are not in the hourly check yet.
What each exchange covers
From the live feed registry- Binance: USDⓈ-M perps and, from 2026-09-28, COIN-M (inverse) perps; liquidations sampled: Binance publishes only the latest liquidation per symbol per 1000 milliseconds
- Bybit: Every liquidation (allLiquidation stream); OI and funding polled every 30s; USDT and inverse perps
- Hyperliquid: No liquidations: Hyperliquid has no public all-market liquidation stream
- Bitget: Sampled liquidations (largest per coin and side each second); OI and funding polled every 30s; no coin-margined perps listed
- Gate: Every liquidation order on USDT perps (futures.public_liquidates, batched each second); USD notional uses the pushed order limit price and can differ from the eventual average-fill notional; OI and funding polled every 30s; BTC-settled inverse perp not covered (under $50M OI)
- HTX: Every liquidation on USDT swaps (public liquidation_orders stream); OI and funding polled every 30s; coin-margined swaps not covered (under $50M OI)
- MEXC: USDT perps: open interest and funding polled over REST every 60s; MEXC publishes no liquidation feed
- KuCoin: USDT and inverse perps: open interest and funding polled over REST every 60s; KuCoin publishes no liquidation feed
- Aster: Crypto perps (stocks excluded): OI every 60s (top 60) or 5 min, funding every 60s; liquidations sampled, at most one per symbol per second. Added 2026-09-28
- Lighter: Every liquidation on crypto perps (liquidation fills on the public trade stream); OI, volume and funding polled every 60s; added 2026-09-28
Open interest: ours vs the exchange's API
median / 95th percentile gap · share within tolerance| Exchange | 1 day | 7 days | 30 days | Tolerance |
|---|---|---|---|---|
| Binance | 0.053% / 0.29% 100% | 0.030% / 0.29% 99.6% | 0.033% / 0.27% 99.8% | 2.00% |
| Bybit | 0.032% / 0.23% 100% | 0.015% / 0.21% 99.3% | 0.015% / 0.18% 99.5% | 1.00% |
| Hyperliquid | 0.070% / 0.32% 99.0% | 0.042% / 0.31% 99.5% | 0.043% / 0.25% 99.7% | 1.00% |
| Gate | 0.051% / 0.22% 100% | 0.038% / 0.34% 99.1% | 0.040% / 0.29% 99.0% | 1.00% |
| Bitget | 0.088% / 0.32% 100% | 0.045% / 0.35% 99.1% | 0.044% / 0.27% 99.3% | 1.00% |
| HTX | — | — | — | 1.00% |
| KuCoin | Not scored: KuCoin's public open-interest endpoints return values from inconsistent caches, often 10-100% apart within seconds, so a gap from them says little about our figure. | 1.50% | ||
| MEXC | Not scored, under review: MEXC's reported open interest moves about 6.8 times more than its own trades imply. Until we know what the figure measures, matching it would only show we copy MEXC's number, not that the number is right. | 1.50% | ||
| Coin totals (sum of venues) | — | — | 100.00% / 100.00% 0.0% | 2.00% |
Funding rate: ours vs the exchange's API
median / 95th percentile gap · share within tolerance| Exchange | 1 day | 7 days | 30 days | Tolerance |
|---|---|---|---|---|
| Binance | 0 bp / 0.01 bp 100% | <0.01 bp / 0.01 bp 99.9% | <0.01 bp / 0.01 bp 99.9% | 0.50 bp |
| Bybit | 0 bp / 0.03 bp 100% | 0 bp / 0.05 bp 99.6% | 0 bp / 0.07 bp 99.3% | 0.50 bp |
| Hyperliquid | <0.01 bp / 0.01 bp 100% | <0.01 bp / 0.01 bp 100% | <0.01 bp / 0.03 bp 100% | 0.50 bp |
| Gate | <0.01 bp / 0.07 bp 100% | <0.01 bp / 0.11 bp 98.9% | <0.01 bp / 0.13 bp 98.5% | 0.50 bp |
| Bitget | 0 bp / 0.01 bp 100% | 0 bp / 0.02 bp 99.9% | 0 bp / 0.01 bp 99.9% | 0.50 bp |
| HTX | — | — | — | 0.50 bp |
| KuCoin | 0 bp / 0.03 bp 100% | <0.01 bp / 0.06 bp 99.4% | <0.01 bp / 0.03 bp 99.8% | 0.50 bp |
| MEXC | — | — | — | 0.50 bp |
Price: ours vs the exchange's API
median / 95th percentile gap · share within tolerance| Exchange | 1 day | 7 days | 30 days | Tolerance |
|---|---|---|---|---|
| Binance | 0.040% / 0.25% 92.8% | 0.033% / 0.21% 94.9% | 0.034% / 0.19% 95.9% | 0.20% |
| Bybit | 0.060% / 0.25% 91.1% | 0.033% / 0.21% 94.7% | 0.033% / 0.23% 94.5% | 0.20% |
| Hyperliquid | 0.078% / 0.27% 85.4% | 0.040% / 0.23% 92.3% | 0.042% / 0.34% 91.5% | 0.20% |
| Gate | 0.041% / 0.24% 91.3% | 0.026% / 0.21% 94.6% | 0.030% / 0.20% 95.1% | 0.20% |
| Bitget | 0.072% / 0.31% 88.0% | 0.040% / 0.37% 91.8% | 0.040% / 0.24% 94.3% | 0.20% |
| HTX | — | — | — | 0.20% |
| KuCoin | 0.073% / 0.29% 85.9% | 0.040% / 0.27% 92.0% | 0.036% / 0.22% 94.2% | 0.20% |
| MEXC | — | — | — | 0.20% |
Stored history, checked nightly
The nightly stored-history check has not run yet.
How fresh each feed is
Last 24 hours · median / 95th percentile| Feed | 1-minute bars | Liquidations | Open interest age | Funding age |
|---|---|---|---|---|
| Aster perps | — / — | — / — | 4.3 min / 9.2 min | 26 s / 26 s |
| Binance perps | — / — | — / — | 3.4 min / 15.2 min | 2 s / 2 s |
| Bitget perps | — / — | — / — | 19 s / 19 s | 19 s / 19 s |
| Bybit perps | — / — | — / — | 4 s / 4 s | 4 s / 4 s |
| Gate perps | — / — | — / — | 16 s / 16 s | 16 s / 16 s |
| HTX perps | — / — | — / — | 15 s / 15 s | 14 s / 14 s |
| Hyperliquid perps | — / — | — / — | 5 s / 22 s | 5 s / 22 s |
| KuCoin perps | — / — | — / — | 42 s / 42 s | 42 s / 42 s |
| Lighter perps | — / — | — / — | 16 s / 16 s | 16 s / 16 s |
| MEXC perps | — / — | — / — | 37 s / 37 s | 37 s / 37 s |
Bars: time from the bar's close to its arrival here. Liquidations: from the exchange's event time to arrival. Open interest and funding age: how long ago each market's value was last refreshed, right now. Exchange clocks can run slightly ahead of ours; those events count as 0 ms.
Last checked: 2026-11-13 14:52 UTC.
Data accuracy FAQ
How accurate is the open interest data?
Every hour we re-fetch a sample of markets' open interest, funding and price from each exchange's own public API and compare them with what this site shows. Over the last 7 days 97.4% of 21,077 scored checks were within tolerance; the tables break that down by exchange. Open interest is always counted one side (longs, which equal shorts). KuCoin and MEXC open interest are not scored: their own figures are unreliable or under review.
Why can a value differ from the exchange's website?
Timing, mostly. Open interest is polled every 30 to 300 seconds depending on the exchange, and bar prices are the last closed 1-minute bar, so a fast market can move past the tolerance between our read and the check. Persistent gaps are what the checks exist to catch.
Where does the data come from?
Only each exchange's own public market-data API. No paid data vendor, no scraping, and no other data site's numbers.
How fast is the data?
The latency table shows, per exchange feed, how long after the exchange's own timestamp a 1-minute bar or a liquidation reached our servers (median and 95th percentile), and how old each market's open interest and funding are right now.