Granular Order Book Updates Now Available in the Market Order Books API

We're rolling out an enhancement to our order books dataset that we think will be valuable for your book-reconstruction and market-microstructure workflows.

What's New

Until now, the /timeseries/market-orderbooks endpoint returned periodic full-depth snapshots of the order book. We've added a new updates mode that delivers incremental, tick-by-tick order book changes alongside snapshots. This lets you reconstruct and continuously maintain a complete, full-depth order book at any arbitrary point in time in the history without repeatedly pulling full snapshots, which substantially reduces data volume and latency for book-reconstruction workflows.

How it Works

Request the new mode with: /timeseries/market-orderbooks?dataset=updates&start_with_snapshot=true

With start_with_snapshot=true, the response opens with a complete full-depth snapshot and is then followed by the stream of incremental updates so you can initialize the full book from the first message and apply each subsequent date.

Each row is tagged with a type of either snapshot (a complete book state) or update (a change to one or more price levels), and carries nanosecond timestamps. Apply rows in the order returned; a price level reported with a size of 0 means that level has been removed.

For real-time use, our /timeseries-stream/market-orderbooks websocket feed delivers the same snapshot-then-updates model live, with a per-message sequence number for ordering.

Initial Coverage

At launch, the updates dataset is available across the following 16 exchanges, spanning spot, futures, and options markets where applicable:

Binance · Coinbase · Coinbase International · Coinbase Derivatives · OKX · Bybit · Kraken · Bitstamp · Gemini · Gate.io · KuCoin · Bitget · Crypto.com · Deribit · CME · Hyperliquid

Historical availability begins in late June 2026. We plan to expand exchange coverage over time; the existing snapshots dataset remains available across our full order book universe in the meantime. We intend to backfill deeper history at a later date, and to additionally deliver the data as Parquet files via Amazon S3.

Additional Information

Please reference our Product Docs for more information on our order book updates dataset where we provide guidance on how to access the data and example responses.

‍