Crypto Historical Data — Free Crypto Dataset Downloads

Download Individual Crypto Datasets

Each ZIP contains the complete timeframe set for one pair: M1, M5, M15, M30, H1, H4, D1, W1 and MN1 — as separate CSV files with the columns Timestamp, Open, High, Low, Close, TickVolume, Spread. Pick your pair below, or open its dedicated page for the full spec sheet, timeframe table and usage guide.

What Is Inside Every Crypto Dataset

• Nine timeframes, one ZIP: M1, M5, M15, M30, H1, H4, D1, W1 and MN1 as separate CSV files.
• Seven columns: Timestamp, Open, High, Low, Close, TickVolume, Spread — broker server time.
• Real broker feed: exported from a live MT5 terminal, not scraped from an exchange API — so it matches what your EA sees in the Strategy Tester.
• 24/7 series: crypto trades weekends, and the data includes those candles — no session gaps to clean.
• 14,036,879 candles across the whole collection, from 15 live pairs and 20 delisted symbols holding their complete frozen history.

M1, M5 and M15 files contain the most recent 100,000 candles each — the MetaTrader 5 terminal window. M30 and above carry the full downloaded history for the symbol.

Why the Date Ranges Differ — Read This Before You Download

We label every dataset honestly instead of pretending all 35 pairs share one range:

Frozen delisted windows are genuinely valuable: the data cannot be revised retroactively, and you can import them into MT5 as custom symbols to reproduce backtests on the exact symbol traders used at the time.

How to Use the Crypto Data

1. Download a pair's ZIP and extract the nine CSV files.
2. For MT5: create a custom symbol (Market Watch → Symbols → Create Custom Symbol → Import Bars) and import the CSVs — this works for delisted symbols too.
3. For Python: load with pandas, parse the Timestamp column, sort ascending and build features (returns, range, volatility, session flags).
4. Validate before modeling: duplicate timestamps, OHLC violations and unexpected gaps are the three checks that catch most bad imports.
5. Keep spread and weekend behavior in your assumptions — crypto CFDs widen at daily rollover and on Saturdays/Sundays.

Crypto Historical Data — FAQ

Is the crypto historical data really free?

Yes — every dataset on this page is free. Clicking a download link opens a one-time email unlock: we send a single confirmation email to verify your address (no spam, never sold or shared). After that your browser is remembered for 30 days and every dataset downloads instantly. Each ZIP contains nine CSV files from M1 to Monthly.

Which crypto pairs are available?

35 datasets: majors like ETH, LTC, BCH and XRP; altcoins like SOL, BNB, ADA, DOGE, LINK and MATIC; bitcoin crosses such as BTCXAU, BTCXAG, BTCJPY, BTCAUD, BTCCNH, BTCTHB and BTCZAR; plus BTCUSDT and ETHBTC.

What date ranges do the crypto datasets cover?

Ranges differ honestly per pair — the table above explains each group. ETH, LTC, BCH and XRP go back to February 2018; BTCJPY to December 2019; most bitcoin crosses and SOL/BNB to 2021; delisted tokens hold their complete history ending at the broker delisting date; BTCUSDT and ETHBTC were listed in 2026.

Why do some datasets end in 2023 or 2024?

Those symbols were delisted by the broker, so the feed stops there. The ZIP contains the complete available broker history for the symbol — frozen windows covering the 2021 bull run and the 2022 bear market. They are ideal for research and custom-symbol backtests, and they cannot be revised retroactively.

Can I use this crypto data for MT5 backtesting and Python?

Yes. The CSV format (Timestamp, Open, High, Low, Close, TickVolume, Spread) imports into MT5 as a custom symbol and loads directly in pandas, Excel, R and machine-learning pipelines such as LSTM, XGBoost and Random Forest. Crypto trades 24/7, so the series contain no session gaps.

More Free Datasets