Market Data API - Historical Tick, OHLCV, Futures, Databento

Pull historical market data from your own Databento account into a dataset, a spreadsheet or a warehouse: tick by tick trades, full order book, top of book, OHLCV bars, instrument definitions and statistics for CME futures and options, US equities, OPRA options, ICE and Eurex, with exact prices and nanosecond timestamps on every row. Uses your own Databento API key, previews the cost before it streams anything, and charges only for rows that carry data.

What it collects

• Market data records from the Databento Historical API in any documented schema: trades, MBO (market by order, L3), MBP-1 and MBP-10 (top of book and 10 levels of depth), TBBO, BBO 1s and 1m, OHLCV bars at 1 second, 1 minute, 1 hour and 1 day, instrument definitions, statistics (settlement, open interest, cleared volume), trading status and auction imbalance.
• Exact prices and timestamps: Databento stores prices as integers in units of 1e-9 and timestamps as UNIX nanoseconds. Each row has the price as a number, the exact nine-place decimal in pricesExact, every timestamp in ISO 8601 with nanoseconds plus a <field>_ns column with the integer, and the documented null sentinels turned into null.
• Cost preview: the cost in US dollars, record count and billable size Databento reports for a request, before anything is streamed. The records mode runs the same check against a ceiling you set and stops if the request would cost more.
• Symbology resolution: raw symbols such as ESM2 or AAPL to instrument IDs and back, parent symbols (ES.FUT) and continuous contracts (ES.c.0), with the date range each mapping holds and the symbols that were not found.
• Dataset metadata: dataset codes with their venues, publishers, the schemas of a dataset, the fields of a schema, the available start and end per schema, the data condition per day (available, degraded, pending, missing) and unit prices in US dollars per GB.
• Read only, pay per result, bring your own key.

How to use it

1. Copy an API key from the API keys page of the Databento portal (32 characters, starting with db-). Paste it into the API key field, which is stored as a secret, or set the DATA_API_KEY environment secret.
2. Run the default Datasets mode to see the dataset codes your key reaches.
3. Run Cost preview with a dataset, symbols, schema and time range. It costs nothing on the Databento side.
4. Run Market data records with the same input. The actor checks the cost against your maximum, then streams up to your maximum results.

Input

Field	What it does
What to read	Datasets (default), publishers, schemas, fields, dataset date range, dataset condition, unit prices, cost preview, market data records or symbology resolution.
Dataset	A dataset code such as GLBX.MDP3, XNAS.ITCH, OPRA.PILLAR or EQUS.MINI.
Symbols	Up to 2,000 symbols, one per line, or ALL_SYMBOLS.
Schema	trades, mbo, mbp-1, mbp-10, tbbo, bbo-1s, bbo-1m, ohlcv-1s, ohlcv-1m, ohlcv-1h, ohlcv-1d, definition, statistics, status or imbalance.
Start, End	YYYY-MM-DD, ISO 8601 such as 2022-06-06T14:30 (UTC unless a zone is given) or UNIX nanoseconds. The end is exclusive.
Input and output symbology type	raw_symbol, instrument_id, parent or continuous.
Field list encoding	Fields mode: json, csv or dbn.
Maximum results	Row cap for the run, sent to Databento as the documented limit in the records mode.
Maximum Databento cost (USD)	Records mode: the run stops before streaming if Databento estimates more. 0 turns the check off.
API key	Your Databento API key, as a secret input, or the DATA_API_KEY secret.

Costs on the Databento side

• Databento bills time series data to your own account, per byte it streams, at the dataset's unit price. New accounts get free credits. Metadata, cost estimates and symbology are free.
• The records mode always sends your maximum results as the documented limit, so Databento stops sending, and billing, at that many records. Repeating the same streaming request is billed again; for large or repeated pulls Databento recommends its batch download in the portal.
• Your Databento license terms for each dataset apply to how you use and redistribute the data. Exchange data such as CME and OPRA usually carries display and redistribution restrictions; this actor delivers data to you only.

FAQ

What is a historical market data API used for?

Getting exchange data at full granularity into your own code. A quant backtests a futures strategy on CME Globex trades and MBP-10 depth. An analyst pulls daily OHLCV bars for a list of US stocks into a spreadsheet. A risk team loads settlement prices and open interest from the statistics schema. A data engineer maps raw symbols to instrument IDs before joining feeds.

Which data source does this actor read?

The Databento Historical HTTP API at hist.databento.com, documented at databento.com/docs under API reference - Historical. It calls the metadata routes (list_publishers, list_datasets, list_schemas, list_fields, list_unit_prices, get_dataset_condition, get_dataset_range, get_record_count, get_billable_size, get_cost), timeseries.get_range in JSON encoding and symbology.resolve. The docs and the live host were checked on 2026-10-05.

Do I need a Databento account?

Yes. This actor is bring-your-own-key and never ships one. Create a Databento account, copy a key from the API keys page and paste it in. A missing or refused key ends the run cleanly with a message saying which it was.

How are prices and timestamps formatted?

Databento prices are fixed-precision integers where one unit is 0.000000001. The actor asks Databento to scale them (pretty_px) and keeps the exact decimal string, for example "4108.500000000", in pricesExact, with the same value as a number in the field itself. Timestamps arrive as ISO 8601 with nine fractional digits (pretty_ts); a ts_event_ns style column carries the exact UNIX nanoseconds. Undefined prices and timestamps are null.

How do I keep the Databento bill under control?

Run the cost preview first, keep the maximum results low while testing, and leave the maximum Databento cost at a value you are happy with. The records mode asks Databento for the cost of the exact request, limit included, and streams nothing if it is higher.

Can I get continuous futures or a whole futures chain?

Yes. Set the input symbology type to continuous for symbols such as ES.c.0 (front month by calendar) or ES.v.0 (by volume), or to parent for ES.FUT (every ES future) or ES.OPT (every ES option).

Can this actor change anything in my Databento account?

No. It reads only and never submits batch jobs. The key is sent only as HTTP basic auth to hist.databento.com, is never logged, and is stripped from any error text.

How is it priced?

Pay per result: one price per market data record, cost preview or metadata row. Requests that find nothing are free. Databento bills the data itself separately to your own account.

Example output

{
  "mode": "records",
  "modeLabel": "Market data records",
  "recordType": "marketData",
  "found": true,
  "dataset": "GLBX.MDP3",
  "schema": "trades",
  "id": "GLBX.MDP3:1",
  "name": "ESM2",
  "symbol": "ESM2",
  "instrumentId": 3403,
  "publisherId": 1,
  "tsEvent": "2022-06-06T00:00:00.070033767Z",
  "tsRecv": "2022-06-06T00:00:00.070314216Z",
  "price": 4108.5,
  "size": 1,
  "side": "A",
  "action": "T",
  "value": 4108.5,
  "rtype": 0,
  "depth": 0,
  "flags": 0,
  "ts_in_delta": 18681,
  "sequence": 157862,
  "ts_recv": "2022-06-06T00:00:00.070314216Z",
  "ts_recv_ns": "1654473600070314216",
  "ts_event": "2022-06-06T00:00:00.070033767Z",
  "ts_event_ns": "1654473600070033767",
  "pricesExact": { "price": "4108.500000000" },
  "retrievedAt": "2026-10-05T09:14:52.118Z",
  "note": null
}

The record is the one in the provider's own get_range example. OHLCV rows fill open, high, low, close and volume; MBP rows carry levels and fill bidPrice, askPrice, bidSize and askSize from the top level; a cost preview fills costUsd, recordCount and billableSizeBytes; a symbol mapping fills symbol, resolvedSymbol, start and end.

Keyword map

market data API, historical market data, tick data API, futures data API, CME data, order book data, level 2 market data, OHLCV bars API, stock market data API, options data OPRA, Databento API, Databento historical, Databento Python alternative, backtesting data.