Matrix series
Every series is a price-by-time matrix, ready to feed a convolutional model without reshaping.
Crypto market microstructure data
Inside
Order book and trade data at price-level resolution. Every trade and every book update from 27 exchanges, aggregated at each price, live and as history from one store.
No account needed for the live preview. What it covers
Simulated data
Finest interval
Every interval is built directly from the raw feed, not resampled.
Exchanges
Binance, Coinbase, OKX, Hyperliquid and 23 more. Spot and perpetuals.
Latency in Europe
From the exchange to your system in Europe over a wired connection, on the shortest intervals.
Five ways to read the market
Merge venues into one series
Every data type can be merged price level by price level across the exchanges, quote currencies and instruments you select, spot and perpetuals together. Nothing enters a merged series that you did not pick. The same request works for history and for the live stream.
Bottom row: the highlighted price level in BTC, per exchange and merged. Simulated data.
What a candle hides
Five patterns a standard candle cannot show. Each one is a column in the data, ready to test.
Absorption
Heavy selling at one price, and price holds. A wick in the candle. A number at a price in the footprint.
One large order or many small ones
The same volume from six prints or four hundred reads differently: one participant, or a sliced order.
Shown versus committed liquidity
A wall that rests all minute and a wall that flashes look identical in a snapshot. The time-weighted book tells them apart.
Market-wide absorption
One venue's book is a fraction of the market. Merged, the footprint shows where the whole market held.
Orders placed and pulled
Size that lands at a price and is pulled before it trades, again and again. The end-of-minute snapshot sees one number. Churn sums every add and cancel in between.
From raw feeds to one series
We maintain the exchange connections, normalise every feed, then aggregate and store the result. Your systems receive one format, whichever exchange the data came from. Time-weighted book and churn are computed from every book update, and no exchange publishes that history after the fact.
27 exchanges
Trades and order books, each in its own format and at its own pace.
CryptickData
Captured, placed on one price grid, aggregated into five data types and merged on request.
Your strategy
The same series over REST for history and WebSocket for live.
Founding partners
Built by Theo Linder and Lily La Morella.
Ten founding places
Founding partners hold the archive from its first candle: two months free, then $2,500 per month fixed for two years, against a standard $5,000. One subscription for the whole company.
Or see live data, no account needed.Built by Theo Linder and Lily La Morella.