QuantRocket

What it is

QuantRocket is a quant research and trading platform that ships as a set of Docker containers you run yourself. JupyterLab is the front end; underneath sit several backtesters rather than one. Zipline is QuantRocket's maintained fork of the engine that powered Quantopian, event-driven and suited to intraday work. Moonshot is a vectorised pandas backtester the vendor wrote itself, with MoonshotML adding walk-forward optimisation over scikit-learn, Keras or XGBoost models. Pipeline does point-in-time screening and ranking across large universes, and Alphalens grades the factors that come out of it. A cron scheduler chains collection, backtest and order submission.

Pricing

There is a free Learning and Evaluation tier — the whole software stack, no credit card, but no live or paper trading and only 2007-2011 US daily prices plus roughly thirty sample symbols.

Everything above it is unpriced in public. The pricing page has a plan card that reads "log in or create account to see pricing", and the underlying price call needs a bearer token. Before you see a number you must register and answer a licence questionnaire that classifies you as professional or non-professional; the default classification is professional, and a professional user who claims otherwise loses the account without a refund. The plan is then sized by number of users and by a maximum account size, billed annually or month-to-month.

Data is the second bill. A paid licence includes QuantRocket's own US stock history and the IBKR borrow-fee and shortable-shares datasets. Sharadar fundamentals, EDI global end-of-day prices and Brain's sentiment datasets are separate purchases, also unpriced in public. Real-time data is a third bill, paid to Interactive Brokers, Alpaca or Polygon directly.

Data & coverage

The included data is end-of-day and historical intraday, not live: US stocks and ETFs back to 2007, delisted names retained. EDI extends that to world exchanges, sold per exchange either as a one-time history purchase or as history with daily updates. Real-time streaming exists as a first-class service, but the feed behind it is always somebody else's subscription.

Integrations

Interactive Brokers and Alpaca for live and paper trading, including IBKR parent-child orders, VWAP algos and combos. Real-time data from IBKR, Alpaca or Polygon. Python client and a CLI over a REST API, so other applications can query data or submit orders without entering JupyterLab.

Limitations

• The price is behind a signup wall. You cannot compare this against anything until you have registered and declared a licence class.
• No hosted option. The licence buys software; the server, its memory and its bill are yours.
• Two brokers, and one of them with a caveat: IBKR Lite has no API, so it must be an IBKR Pro account.
• The backtesters cover equities, futures and FX. Options appear only as real-time data and live orders through IBKR — there is no options backtesting.
• Data is licensed for internal use only. No redistribution, EDI data cannot leave QuantRocket, and the included datasets stop working when the licence does.
• The last numbered release is 2.11.0, dated 28 January 2025, and the blog stopped in December 2024. What continues is maintenance rather than features: the containers are still being patched (the realtime service in April 2026, the IB Gateway service in August 2026) and the founder was still answering the support forum in August 2026, while the Python repositories on GitHub last moved in 2024 — quantrocket-client in November 2024, Moonshot in August 2024.

Alternatives

QuantConnect and its LEAN engine for the hosted equivalent with more brokers and published prices; zipline-reloaded, backtrader, VectorBT or Nautilus Trader if you want the engine without the platform and are willing to assemble the data yourself.