Download forex historical data
without rebuilding the pipeline
HistoricalFX packages audited OHLCV forex data into Parquet files for Python, pandas, DuckDB, and local research workflows. Every paid release includes coverage reporting so known gaps are visible instead of hidden.
Current audited release
Why this exists
The raw price feed is only part of the job. The commercial value is the time saved between finding source files and having something defensible enough to test.
Free sources still require downloading, cleaning, de-duplicating, and converting before a backtest can run.
Broker history centers hide data limitations until a strategy test fails or produces suspicious results.
Raw CSV files are slow to load, easy to parse incorrectly, and hard to validate across many symbols.
Gap visibility matters: hidden missing bars are more dangerous than documented coverage limitations.
Best first download
Start with the major-pair audit bundle if you want the most practical first purchase. These pairs currently form the strongest first-dollar surface while deeper source-backed backfill continues.
Get Major Pair BundlePair coverage page, audit facts, and download CTA.
Pair coverage page, audit facts, and download CTA.
Pair coverage page, audit facts, and download CTA.
Pair coverage page, audit facts, and download CTA.
Pair coverage page, audit facts, and download CTA.
See exactly what is sellable now and what is still being backfilled.
Try before you buy
Get a free EUR/USD M1 Parquet sample: 31,680 rows from January 2024, with no nulls or duplicate timestamps.
No spam, ever. Just data.