PyBroker

PyBroker is a Python backtesting framework for strategies that include machine-learning models. It supports custom data sources, model training, strategy optimization, and walk-forward analysis in the same research workflow.

machine learningstrategy backtestingstrategy optimizationPython framework
GitHub Stars3,463
Primary languagePython
LicenseApache-2.0 with Commons Clause
Official repositoryOpen GitHub ↗

Is it a fit?

The quick decision view, grounded in the ranking data and verification record.

Best for

People evaluating the documented machine learning, strategy backtesting use case.

Not ideal for

Anyone who needs capabilities beyond what the official project currently confirms.

Tested status

Not locally tested

Setup difficulty

Low

Last verified

2026-08-02

What it can do

Research and testing

Strategy backtesting
Yes
LLM or agent based
No

Trading

Paper trading
Not confirmed
Automated trading
Not confirmed

Getting started

API key required to start
No
Registration required to start
No
Self-hostable
Yes

Capabilities shown as Yes or No come from the official project source. Unconfirmed fields remain “Not confirmed”.

Related task rankings

The same tool can rank differently for different jobs.

Not currently ranked in the open-source track.