freqtrade

TLDR

Start trading with a configuration and strategy

$ freqtrade trade --config [config.json] --strategy [StrategyName]

Run backtesting on historical data

$ freqtrade backtesting --config [config.json] --strategy [StrategyName]

Download historical data for backtesting

$ freqtrade download-data --config [config.json] --pairs [BTC/USDT] --timeframes [1h]

Create a new configuration file interactively

$ freqtrade new-config

Optimize strategy parameters with hyperopt

$ freqtrade hyperopt --config [config.json] --strategy [StrategyName] --hyperopt-loss [SharpeHyperOptLoss]

List available exchanges

$ freqtrade list-exchanges

List trading pairs on an exchange

$ freqtrade list-pairs --config [config.json] --exchange [binance]

Start the web UI

$ freqtrade webserver --config [config.json]

SYNOPSIS

freqtrade command [options]
freqtrade trade [-c config] [-s strategy]
freqtrade backtesting [-c config] [-s strategy]

freqtrade is an open-source cryptocurrency trading bot written in Python. It supports automated trading on major exchanges including Binance, Kraken, Coinbase, and many others through the CCXT library.
The bot executes user-defined trading strategies that specify entry and exit conditions based on technical indicators. Strategies are Python classes that can be backtested against historical data before live deployment. The hyperopt feature uses machine learning to optimize strategy parameters.
Freqtrade operates in dry-run mode by default, simulating trades without real money. It can be controlled via Telegram bot or the FreqUI web interface. Trade data is stored in a SQLite database for analysis and the plot commands visualize strategy performance.

PARAMETERS

-c, --config PATH

Specify configuration file (default: config.json)

-s, --strategy NAME

Specify strategy class name

--strategy-path PATH

Additional strategy lookup path

-d, --datadir PATH

Path to historical data directory

--userdir PATH

Path to user data directory

-v, --verbose

Verbose mode (-vv for more, -vvv for all)

--logfile FILE

Log to specified file

--no-color

Disable colorized output

--db-url URL

Database URL for trade storage

--dry-run

Run in simulation mode without real trades

-V, --version

Show version number

-h, --help

Show help message

COMMANDS

trade

Start live or dry-run trading

backtesting

Test strategy on historical data

hyperopt

Optimize strategy parameters

download-data

Download historical OHLCV data

new-config

Create new configuration interactively

create-userdir

Create user data directory structure

list-exchanges

Print available exchanges

list-pairs

Print available trading pairs

list-strategies

Print available strategies

list-timeframes

Print available timeframes

show-trades

Display trades from database

plot-dataframe

Plot strategy signals on chart

plot-profit

Plot profit over time

webserver

Start the FreqUI web interface

install-ui

Install the FreqUI web interface

CAVEATS

Cryptocurrency trading carries significant financial risk. Backtesting results do not guarantee future performance. Exchange APIs may have rate limits that affect bot operation. Strategies require thorough testing in dry-run mode before live trading. Configuration files contain API keys and should be secured properly. The bot requires continuous operation and stable network connectivity.

HISTORY

Freqtrade was started in 2017 as a community-driven open-source project. Written in Python 3, it grew from a simple trading bot to a comprehensive trading framework with backtesting, optimization, and visualization tools. The project gained significant traction during the 2020-2021 cryptocurrency bull market. It introduced FreqAI in 2022 for machine learning-based adaptive strategies. Development continues actively on GitHub with regular releases.