Historical data

How to access OHLCV history, trade ticks, and order book snapshots for backtesting and research

Kraken’s public REST endpoints provide historical market data without authentication. This page covers what’s available, the limits per call, and how to paginate efficiently for bulk collection.

Available data types

Data type	Endpoint	Depth	Authenticated
OHLCV candles	GET /0/public/OHLC	Up to 720 candles per call	No
Trade ticks	GET /0/public/Trades	Up to 1000 trades per call	No
Order book snapshot	GET /0/public/Depth	Up to 500 levels per side	No
Grouped book	GET /0/public/GroupedBook	Configurable tick size	No
L3 order book	POST /0/private/Level3	Full book	Yes

OHLCV candles

The OHLC endpoint returns candlestick data for a given interval.

GET /0/public/OHLC?pair=XBTUSD&interval=1440&since=1700000000

Parameters:

Parameter	Values	Default
pair	Instrument name (e.g. XBTUSD)	Required
interval	1, 5, 15, 30, 60, 240, 1440, 10080, 21600 (minutes)	1
since	Unix timestamp — return candles after this time	Oldest available

Response structure:

{
  "result": {
    "XXBTZUSD": [
      [1700000000, "37000.0", "37500.0", "36800.0", "37200.0", "37100.0", "150.5", 320],
      ...
    ],
    "last": 1700086400
  }
}

Each candle: [time, open, high, low, close, vwap, volume, count]

Pagination: use the last value from the response as the since parameter in your next call to page forward.

Depth limits:

• Maximum 720 candles per call
• Daily candles (interval=1440): ~720 days (~2 years) per call
• Minute candles (interval=1): 720 minutes (~12 hours) per call

Practical tip for collecting a year of daily data:

import requests, time

def fetch_ohlcv(pair, interval, since=None):
    params = {"pair": pair, "interval": interval}
    if since:
        params["since"] = since
    r = requests.get("https://api.kraken.com/0/public/OHLC", params=params)
    data = r.json()
    result = list(data["result"].values())[0]  # candles
    last = data["result"]["last"]
    return result, last

all_candles = []
since = None

while True:
    candles, last = fetch_ohlcv("XBTUSD", 1440, since)
    if not candles:
        break
    all_candles.extend(candles)
    if since == last:
        break
    since = last
    time.sleep(1)  # respect rate limits

Trade ticks

The Trades endpoint returns a time-ordered list of public trades.

GET /0/public/Trades?pair=XBTUSD&since=1700000000000000000&count=1000

Parameters:

Parameter	Description
pair	Instrument name
since	Nanosecond timestamp — return trades after this time
count	Max trades per call (up to 1000)

Response structure:

{
  "result": {
    "XXBTZUSD": [
      ["37200.00000", "0.05000000", 1700000001.1234, "b", "l", "", 12345678],
      ...
    ],
    "last": "1700000100000000000"
  }
}

Each trade: [price, volume, time, buy/sell, market/limit, miscellaneous, trade_id]

Pagination: use last as the since value in the next call. The since value is in nanoseconds.

Rate limit note: the Trades endpoint is public and subject to IP-based rate limits. For bulk historical collection, introduce a delay between requests (100–200ms) to avoid throttling.

Order book snapshots

The Depth endpoint returns the current order book at the time of the request. It is not a historical endpoint — it reflects the live state.

GET /0/public/Depth?pair=XBTUSD&count=100

For L3 (order-by-order) book data, use the authenticated Level3 endpoint:

POST /0/private/Level3
# body: pair=XBT/USD

L3 data includes individual OrderID values, enabling reconciliation with ExecutionReport data from FIX or the executions WebSocket channel.

Futures historical data

Futures provides dedicated historical endpoints under the Futures REST API:

Data	Endpoint
OHLCV candles	GET /api/charts/v1/{tick_type}/{symbol}/{resolution}
Trade history (fills)	GET /derivatives/api/v3/history
Historical funding rates	GET /derivatives/api/v3/historicalfundingrates

# Get BTC perpetual daily candles
GET https://futures.kraken.com/api/charts/v1/trade/PI_XBTUSD/1D

# Get historical funding rates
GET https://futures.kraken.com/derivatives/api/v3/historicalfundingrates?symbol=PI_XBTUSD

Building a backtesting dataset

A practical approach for collecting a clean historical dataset:

1. Collect OHLCV for your timeframe

   Start with daily candles to get a broad picture, then drill into minute candles for the periods you need. Use the pagination pattern above.

2. Augment with trade ticks for microstructure

   Trade ticks give you the actual execution prices and sizes. Useful for modelling slippage and for tick-by-tick backtesting.

3. Validate against VWAP

   Each OHLC candle includes VWAP. Cross-check your collected trade ticks against the VWAP to verify completeness and catch any pagination gaps.

4. Handle gaps

   Thin markets or API collection gaps can leave missing candles. Decide upfront how to handle them: forward-fill with the previous close, or exclude those periods from your strategy.

Kraken does not provide a bulk historical data dump or websocket replay service. For very deep history (years of tick data) consider supplementing with third-party data providers.

Rate limits for data collection

Historical data endpoints are public but rate-limited per IP. Recommended safe intervals:

Endpoint	Safe delay between calls
OHLC	1 second
Trades	1–2 seconds
Depth	1 second

If you need faster collection, distribute requests across multiple IPs or use the FIX market data feed for real-time capture going forward.

Related guides

API comparison

FIX and WebSocket for real-time market data capture

Rate limits

IP-based rate limits for public data endpoints

WebSocket introduction

Stream live market data for forward capture alongside historical backfill