Coin Metrics is pleased to announce the version 2.10 release of our CM Market Data Feed. This release contains support for new exchanges, expanded coverage for many of our data types, significant quality-of-life improvements to our API, and many more upgrades and bug fixes.

What’s New

  • New Exchanges: Coin Metrics is constantly expanding our exchange coverage or adding additional data types for exchanges we already cover. This release contains the following new exchanges and their supported data types listed below.
    • ErisX (Cboe Digital) spot markets metadata, trades, real-time order books, full depth order book snapshots, and candles. Due to our agreement with Cboe Digital, we do not permission users by default, but if you are interested in accessing the data, please reach out to us.
    • Deribit spot markets metadata, trades, candles, and volume metrics.
    • Binance aggregated futures markets metadata, trades, and candles. Binance serves each individual trade for their futures markets through their HTTP API endpoint only (not websocket API). Binance instead serves trades where individual trades with the same price and side in 100 millisecond intervals are aggregated to one observation through their websocket API. Coin Metrics previously solely relied on Binance’s HTTP API endpoint to collect every individual, non-aggregated trade. However, this collection method introduced a small amount of latency. If users are interested in Binance futures trades with the lowest possible latency and prefer to work with aggregated trades data, we now offer a new exchange binance_aggregate for this purpose.
    • BitMEX futures markets real-time order books and full depth order book snapshots.
    • Bithumb spot markets metadata, trades, and candles.
    • Bullish spot markets trades history was backfilled to November 2021, the inception of the exchange. Candles and volume metrics were also backfilled.
  • Expansion of our asset coverage for CM Prices and inclusion of defi markets in our pricing. We released a new version of our CM Prices which brings our coverage universe for the CM Reference Rates and CM Principal Market Prices to 887 assets. With our latest release, we now have the capability to include defi markets as constituents in the calculation of our prices which allows us to expand our coverage universe further and improve the quality of existing assets that trade primarily on defi markets.
  • Expanded the universe of assets and pairs for aggregated quotes. Our aggregated quotes are analogous to national best bid-offer for equities and are served through our /timeseries-stream/asset-quotes and /timeseries-stream/pair-quotes endpoints. We now support 45 of the top assets and support for all pairs in our coverage universe.
  • Enhanced completeness for candles and volume metrics. Coin Metrics calculates candles and volume metrics based on trades that we collect from exchanges. In rare circumstances, trades are collected with a slight lag which in the past may have caused our candles and volume metrics data to be incomplete. Now, we have enabled automatic detection of late trades and will automatically recalculate our candles and volume metrics so that our data will include all trades, even ones collected with a lag.
  • New constituent endpoints. Many of our users are interested in the constituents that are used in the calculation of our data, such as which markets are used in our CM Prices or which exchanges are included in our trusted volume metrics. To address this need, we have created new /constituent-snapshots/asset-metrics and /constituent-timeframes/asset-metrics API endpoints to serve the constituents for our data. Currently, only our trusted volume metric is supported, but we hope to add support for more metrics over time.
  • Real-time open interest and liquidations data. We now serve our open interest and liquidations data through our websocket API endpoints /timeseries-stream/market-openinterest and /timeseries-stream/market-liquidations for users who prefer to consume this data with the lowest possible latency. Previously, we only served this data through our HTTP endpoints.
  • Daily candles with arbitrary ending time. Users have the ability to query daily candles that end at any arbitrary hourly timestamp by using the frequency value 1d-HH:00, where HH can take values from 00 to 23. This option is supported for both /timeseries/market-candles and /timeseries/pair-candles. Previously, we only supported daily candles that ended at midnight UTC time.
  • Contract prices for futures markets. We now offer the index and mark prices for futures markets through /timeseries/market-contract-prices. Currently, Deribit and Bybit futures markets are supported and more exchanges will be supported in the future.
  • Support for the csv format for all endpoints. All of our endpoints which return timeseries data now support csv as a value for our format parameter, as well as json which is returned by default.
  • Options to force the API to return all available data even when errors are encountered. Previously, our API would return an error if a user requested data that is not supported or that they are not permissioned to. Our API would not return any data in this circumstance. We have added new parameters ignore_unsupported_errors and ignore_forbidden_errors that will force the API to return all data queried by the user that is supported or that the user is permissioned to, respectively, even if the user request would normally result in an error. These new parameters are supported for /timeseries/asset-metrics and are useful for interactive use or for adding resilience in scheduled jobs.
  • Expanded wildcard support. Our API now supports wildcards for many more endpoints which allows a user to query our entire coverage universe without specifying each individual entity. Previously, wildcard support was limited to certain endpoints and now it has expanded to include more endpoints. For the full list, please consult our API4 documentation.
  • Added a multiplier field for /reference-data/markets. This field is relevant to understanding the futures specifications for certain futures markets on some exchanges.

Other Announcements

  • Trusted Exchange Framework v2.1. We published our Trusted Exchange Framework v2.1 which contains improvements on how we evaluate exchanges in the regulatory category and a new letter grading scale. Coin Metrics utilizes the Trusted Exchange Framework as a systematic and rules-based method to select constituents for our prices, indexes, and metrics. Our CM Prices and trusted volume metrics were updated to use the latest trusted exchanges.
  • Coin Metrics Coverage now has expanded support for more data types. Our Coin Metrics Coverage website now includes all our market data types and pulls data from our catalog v2 endpoints. Users can easily explore our full coverage universe using this tool.


  • We improved our trades and order book data collection latency for all exchanges, especially Coinbase and Binance.
  • Several optimizations to improve the stability of our connection and to reduce our data collection latency with Bybit.
  • We improved our collection of OKEx spot markets to ensure that our order book snapshots will always contain the maximum levels supported by the exchange or all levels within 10 percent from the midprice, whichever is greater.
  • Significant performance improvements in our API, especially in reducing the latency for querying our order book snapshots and metrics served through our HTTP endpoints, as well as our real-time trades and order book data served through our websocket API endpoints.

Bug and Data Quality Fixes

  • Fixed a bug that prevented Gemini’s order book snapshots from being sorted by price.
  • Recollected Bitfinex’s liquidations data to make it consistent with how other exchange’s report liquidations data and prevent double-counting of liquidations. Bitfinex’s liquidation metrics were also recalculated to remove the impact of double-counting. More information can be found in our status page incident report.
  • Fixed a bug that would cause us to not collect trades from Kraken under rare circumstances.
  • Fixed an issue with our Binance futures candles on 2023-11-09 that resulted in incomplete data and recalculated downstream volume metrics.
  • Fixed some missing metadata and ticker conflicts in order books snapshots.
  • Fixed an issue with zero low price in pair candles on 2023-08-13 to 2023-08-26.
  • Backfilled missing order book data for some exchanges from 2022-09-12 to 2022-12-09.
  • Fixed an issue with missing CME futures open interest data from 2023-09-15 to 2023-10-05 and recalculated downstream open interest and funding rate metrics.