lumibot

What it is

Lumibot is a Python framework to build, backtest, and run algorithmic trading strategies and AI agents in Python. It supports deterministic strategies and AI-agent strategies, backtests that replay historical data, and paper or live trading using real brokers.

How it works

The project provides a strategy base, backtesting modules (e.g., YahooDataBacktesting), and broker integrations (e.g., Alpaca). It includes built-in AI agent tools (market state, filings, indicators, memory, and notifications) and supports running agents within the normal Lumibot strategy loop for either backtesting or live trading. The README shows usage patterns for backtesting and running with a broker, and includes a sample multi-agent setup:

• Researchers, bull/bear agents, and a trader agent that can submit orders through Lumibot.
• Agents can access market data, SEC filings, macro data, indicators, and local memory.

Getting started

pip install lumibot

from lumibot.strategies import Strategy
from lumibot.backtesting import YahooDataBacktesting
class MyStrategy(Strategy):
    def on_trading_iteration(self):
        if self.first_iteration:
            aapl = self.create_order("AAPL", 10, "buy")
            self.submit_order(aapl)
MyStrategy.backtest(
    YahooDataBacktesting,
    datetime(2023, 1, 1),
    datetime(2024, 1, 1),
)

export ALPACA_API_KEY='your-alpaca-key'
export ALPACA_API_SECRET='your-alpaca-secret'
export ALPACA_IS_PAPER=true
python my_strategy.py

import os
from lumibot.brokers import Alpaca
from lumibot.traders import Trader
ALPACA_CONFIG = {
    "API_KEY": os.environ["ALPACA_API_KEY"],
    "API_SECRET": os.environ["ALPACA_API_SECRET"],
    "PAPER": os.environ.get("ALPACA_IS_PAPER", "true").lower() != "false",
}
broker = Alpaca(ALPACA_CONFIG)
strategy = MyStrategy(broker=broker)
trader = Trader()
trader.add_strategy(strategy)
trader.run_all()
```

## Recent releases
Latest releases include v4.5.83 (2026-08-05) with AI-native options trading tools; prior 4.5.82 (2026-08-04) focused on state bounds and cache self-healing; 4.5.81 (2026-07-30) includes IBKR cache gap self-healing; 4.5.80 (2026-07-30) and 4.5.79 (2026-07-30) included broker-neutral data reliability and backtest improvements.

## Traction
stars_7d not provided; stars_ic not provided in FACTS. Raw numbers present: 1883 stars, 362 forks.

## Behind the repo
Lumibot is developed under Lumiwealth and linked to BotSpot services for managed cloud deployment, documentation, and a marketplace for AI coding agents. The README references BotSpot for deployment and notes AI agent capabilities and integration with various data sources.

## Caveats
License: GPL-3.0. Created 2020-09-10. Last push 2026-08-04. Open issues: 85. Language: Python. Topics include ai-agents, algorithmic-trading, alpaca, backtesting, broker, crypto-trading, interactive-brokers, llm-agents, and SEC data.