What is QuantLab?
QuantLab is an MCP server you add to your AI coding environment (Cursor, Claude Code, and similar clients). Once it is connected, you can ask for quantitative equity research in ordinary language: pull market data, explore ideas, run backtests and stress tests, compare portfolio constructions, and look at performance and risk metrics — without assembling a separate research stack for each step. Details on how to attach the server live in Quickstart. From there, Example prompts show how people typically talk to it.Why use it
One place for the loop
Data, experiments, backtests, stress scenarios, and reporting stay in the same conversation instead of jumping across disconnected tools.
See the data yourself
Download market history, open the results, and inspect charts and numbers with your own eyes — not only a model’s summary.
Iterate freely
Change tickers, dates, methods, and hyperparameters; rerun what matters. Independent experiments can run in parallel when you want to compare many setups at once.
From research to something you can ship
Combine QuantLab runs with your local code and notebooks, lock in what works, and carry the strategy forward toward deployment.