Qlib

Quant research platform for modelling, signal building and backtesting

🧩 Build it Free 02 in Backtesting 🆕 NEW TODAY

StocksStandalonePython

48,442 MITupdated 2 Sep '26

Overview

Qlib is an AI-oriented quantitative investment platform covering data processing, model training, backtesting, and order execution, built around YAML-configured workflows.

Forquant researchers who want a full pipeline (data, model zoo, backtest) rather than a single-purpose backtester.
NeedsPython 3.8-3.12 (conda recommended), and a downloaded data bundle before any workflow will run.
Runsa Python package (pip install pyqlib) with YAML workflow files run through its qrun CLI.
Limitthe official data bundle is temporarily disabled under Microsoft's own data security policy; the README points to a community mirror instead.

In guides

The 10 best GitHub repos for vibe-coding a trading tool

Vibe it

Paste this into your AI agent. Write the first line, or take a suggestion; the copy carries your words.

Source: https://github.com/microsoft/qlib — clone it and read its README and examples before writing anything; its documented setup beats what you remember.
Stack: a Python quant research platform — a data layer, a model zoo (supervised, market-dynamics, reinforcement learning), and a backtest layer wired together by YAML workflow files.
Needs: Python 3.8-3.12 (conda recommended), and a downloaded data bundle before any workflow will run.
Start: `pip install pyqlib`, then download data with `python -m qlib.cli.data qlib_data --target_dir ~/.qlib/qlib_data/cn_data --region cn` (the official dataset is currently disabled — the README points to a community mirror) before running qrun on a bundled workflow YAML.
Limit: the official dataset is temporarily disabled under Microsoft's own data security policy; only a community-contributed mirror is available for now.
Done means: you show me the report output from one bundled workflow YAML run through qrun, and confirm which data source you used.
Before you start, ask me: whether I want the CN A-share default data or my own data source, and which bundled model or workflow YAML to start from.

More in Backtesting

All 6 →

Backtrader 🧩 Build it Python backtesting and live-trading library with built-in plotting

23,167

Backtesting.py 🧩 Build it Small Python framework for backtesting strategies on candlestick data

8,949

PyAlgoTrade 🧩 Build it Event-driven Python library for backtesting, paper and live trading

4,675

Something wrong here? Tell us →