Metrics V2

Overview

The Metrics V2 data type provides 1-minute aggregates of CoinAPI Metrics V2 observations for each exchange. Each row is one metric series over a UTC minute and includes the first, last, minimum, maximum, and sum of values in that window, along with the observation count.

Use Metrics V2 data to:

• Backfill historical mark price, index price, and estimated delivery price
• Reconstruct funding-rate and open-interest time series without polling REST endpoints
• Analyze liquidation activity and 24-hour ticker statistics at minute resolution
• Join metric bars with trades, quotes, and OHLCV using CoinAPI symbol identifiers

Normalized coinapi_metric_id values match the Metrics V2 listing and the Metric ID per Exchange reference.

[!NOTE]

This dataset is available only in the coinapi bucket. It is not included in coinapi-daily-tail.

File Organization

Metrics V2 files are stored only in the coinapi bucket. Current files are published as hourly partitions. Some earlier dates also appear as daily D-YYYYMMDD prefixes.

T-METRICS-V2/
└── D-YYYYMMDDHH/
    └── E-[EXCHANGE]/
        └── TP-1MIN.csv.gz

Example:

T-METRICS-V2/D-2026091013/E-BINANCEFTS/TP-1MIN.csv.gz

Where:

Each file contains all metric series for the exchange in that hour, one row per metric, symbol (or asset/chain), and minute.

A Parquet artifact (one file per exchange per hour) is also available at:

T-METRICS-V2/D-YYYYMMDDHH/E-[EXCHANGE].parquet

File Format

Files are encoded as CSV using semicolon (;) as the delimiter and gzip compression. The header row is always present. Each row represents one 1-minute aggregate for a single metric series.

Data Fields

coinapi_metric_id is the field to filter on for normalized analytics. exchange_metric_id preserves the original exchange source.

Example Data

Below is a sample of 1-minute Metrics V2 bars for BINANCEFTS during hour 2026-09-10 13:00 UTC.

asset_id;exchange_id;exchange_metric_id;coinapi_metric_id;exchange_symbol_id;coinapi_symbol_id;chain_id;network_id;time;first;last;min;max;sum;count;minTime;maxTime
;BINANCEFTS;<symbol>@markPrice.P;DERIVATIVES_MARK_PRICE;BTCUSDT;BINANCEFTS_PERP_BTC_USDT;;;2026-09-10T13:15:00.000;76955.53387065;76946.52298463;76946.52298463;76955.53387065;1539019.8992676297;20;2026-09-10T13:15:00.265;2026-09-10T13:15:57.180
;BINANCEFTS;<symbol>@markPrice.r;DERIVATIVES_FUNDING_RATE_CURRENT;BTCUSDT;BINANCEFTS_PERP_BTC_USDT;;;2026-09-10T13:00:00.000;7.952E-05;7.952E-05;7.952E-05;7.952E-05;7.952E-05;1;2026-09-10T13:00:08.932;2026-09-10T13:00:08.932
;BINANCEFTS;/fapi/v1/openInterest.openInterest;DERIVATIVES_OPEN_INTEREST;BTCUSDT;BINANCEFTS_PERP_BTC_USDT;;;2026-09-10T13:21:00.000;108094.088;108094.088;108094.088;108094.088;108094.088;1;2026-09-10T13:21:10.831;2026-09-10T13:21:10.831
...

Data Collection Process

1. We collect Metrics V2 observations from supported exchange streams and REST endpoints.
2. Values are normalized to CoinAPI metric, exchange, and symbol identifiers.
3. Observations are aggregated into 1-minute bars (first, last, min, max, sum, count).
4. Data is uploaded to the coinapi bucket as hourly partitions (D-YYYYMMDDHH), with one TP-1MIN.csv.gz file per exchange.

Corner Cases and Special Considerations

1. Sparse metrics – Low-frequency series such as funding rate or open interest often have count = 1, so first, last, min, max, and sum are identical.
2. Empty identifiers – asset_id, chain_id, and network_id are empty for exchange-symbol metrics. coinapi_symbol_id can also be empty when an exchange symbol is not yet mapped.
3. Non-price values – Some metrics store timestamps or other numeric encodings (for example DERIVATIVES_UPDATE_TIME, DERIVATIVES_FUNDING_TIME). Treat first / last / min / max according to the metric type, not as prices.
4. Time alignment – time is the UTC minute boundary. minTime and maxTime are the actual observation timestamps inside that minute.
5. Coverage – Metric availability varies by exchange. Use the Metrics V2 listing endpoints to see which coinapi_metric_id values exist for a given exchange or symbol.

Usage Tips

1. Filter by coinapi_metric_id (for example DERIVATIVES_MARK_PRICE) before joining to other datasets.
2. Join coinapi_symbol_id to Market Data REST symbol metadata (symbol_id) when you need instrument attributes.
3. Use last as the minute's latest value for mark price, funding rate, or open interest charts; use min / max for intraperiod range.
4. Download a single TP-1MIN.csv.gz file per exchange-hour, then filter locally by symbol or metric instead of expecting per-symbol object keys.

For questions or issues regarding Metrics V2 data, please reach out to the support team.