LumiBot vs freqtrade¶
freqtrade and LumiBot are both free, open-source Python trading frameworks with backtesting and live trading. The biggest difference is scope. freqtrade is built for crypto exchanges. LumiBot covers stocks, options, futures, forex, crypto and prediction markets, and lets AI agents make decisions inside the strategy.
Both are good projects. Pick the one that fits the markets you trade and how you want to build.
At a glance¶
freqtrade |
LumiBot |
|
|---|---|---|
What it is |
Open-source crypto trading bot, controlled through Telegram or a web UI |
Open-source Python framework for trading strategies and AI trading agents |
Language |
Python 3.12+ |
Python 3.10+ |
License |
GPL-3.0 |
GPL-3.0 |
Assets |
Crypto: spot, plus futures and leverage on supported exchanges |
Stocks, ETFs, options, futures, forex, crypto and prediction markets |
Live trading |
Yes. Spot on Binance, BingX, Bitget, Bybit, Gate, HTX, Hyperliquid, Kraken and OKX; futures on Binance, Bitget, Bybit, Gate, Hyperliquid, Kraken and OKX. Other exchanges may work through CCXT. |
Yes. Alpaca, Interactive Brokers, Tradier, Schwab, Tradovate, ProjectX, Bitunix, Polymarket and many crypto exchanges through CCXT. See Trading Brokers Supported by LumiBot. |
Backtesting |
Yes, plus hyperopt parameter search, lookahead analysis and recursive analysis |
Yes, with charts, trade files, logs and tearsheets. See Backtesting Trading Strategies in Python with LumiBot. |
AI and machine learning |
FreqAI, adaptive machine learning models that retrain on market data |
Built-in LLM agents that research, call tools and place orders inside the strategy. See Build AI Trading Agents in Python with LumiBot. |
Dry run / paper |
Dry-run mode |
Paper trading through supported brokers |
Hosted option |
Self-hosted |
Self-hosted, or BotSpot for hosted backtests and deployment |
When to choose freqtrade¶
Choose freqtrade if you only trade crypto and want a mature bot you can run and control from Telegram or a web UI. Its hyperopt tools and FreqAI machine learning module are strong for tuning crypto strategies. It has a very large community.
When to choose LumiBot¶
Choose LumiBot if you trade more than crypto, for example stocks, options or futures, or want one strategy style across several brokers. Choose it if you want large language model agents to research news, filings or prices and decide trades inside a backtest you can inspect. Strategies are normal Python classes, and hard rules such as position limits and order sizing can stay in Python.
Moving from freqtrade to LumiBot¶
A freqtrade strategy works on a dataframe of candles and sets entry and exit
signals. In LumiBot you write a Strategy class and place orders in
on_trading_iteration. You can still use pandas and indicator libraries.
Start with Start with LumiBot and the crypto broker list in
Crypto Exchanges for LumiBot.
Try LumiBot with your AI assistant¶
The fastest way to try LumiBot is to let your AI coding assistant do the setup. Install the free LumiBot trading skills in Claude Code, Codex or Cursor with one command, then connect the free BotSpot MCP so your assistant can run backtests in the cloud. See Use LumiBot with Claude Code, Codex and Cursor for the exact commands.
Strongly recommended: connect the free BotSpot MCP
Use Claude, ChatGPT, Codex or Cursor? Connect the free BotSpot MCP and your AI gets:
Strategy-building skills for LumiBot.
Free data: news, economic data and SEC filings.
Backtests with charts it can run and read for you.
Cheap scheduled deployments on BotSpot’s servers, so you do not have to keep your own computer or server running.
The community and marketplace: see what other people’s bots are doing, post your own, research strategies and publish yours.
Easy broker linking.
The BotSpot MCP is free to connect. For Claude, click Add custom connector
and paste https://mcp.botspot.trade/mcp.
It also gives LumiBot’s own AI agents extra research tools inside your strategies. Connect it free at botspot.trade/agents · Why connect it
Sources¶
Facts on this page were checked on October 24, 2026, against each project’s own repository. Neither project guarantees profitable trading; backtests are simulations.