FinRL
Financial reinforcement learning framework for trading strategies.
Build, train, and backtest deep reinforcement learning trading agents.
FinRL is an open-source framework designed for financial reinforcement learning and quantitative trading workflows. It suits researchers, developers, and students wanting to prototype train-test-trade pipelines.
What it is
The framework organizes operations into market environments, DRL agents, and financial applications using a train-test-trade pipeline. It supports algorithms from Stable Baselines 3, RLlib, and ElegantRL, integrating with multiple data processors.
Who it's for
It suits quants, researchers, and developers exploring algorithmic trading and reinforcement learning. It does not suit those seeking out-of-the-box production live trading without migrating to FinRL-X.
Key features
Three-Layer Architecture
Separates market environments, DRL agents, and financial applications for modular design.
Multiple DRL Backends
Supports training via Stable Baselines 3, RLlib, and ElegantRL algorithm libraries.
Technical Indicators
Includes built-in preprocessors for MACD, RSI, VIX, and turbulence index calculations.
Train-Test Pipeline
Provides explicit workflow scripts for downloading data, training agents, and backtesting.
Up and running in 3 steps
Our honest verdict
What's great
- Original pioneer framework for financial reinforcement learning
- Integrates with multiple established DRL libraries
- Provides sample datasets and stock trading tutorials
- Modular structure dividing meta environments and agents
Where it falls short
- Original repository is designated for education while production requires FinRL-X
- Requires manual virtual environment and python dependency setup
- Depends on external data source availability like Yahoo Finance
Pricing & license
Free under the MIT license, with the only cost being your own compute resources for training models.
Open remote AI-training roles
880 open remote AI-training roles, updated November 15, 2026; most pay $20–$175 an hour. Zekai lists them free; the application goes to the employer’s platform.
Alternatives
These reviewed AI tools do the same job instantly — from our scored directory.
Use it, or have us set it up
FAQ
Is this the latest version of FinRL?
No, this repository preserves the original educational framework, while production-oriented development has moved to FinRL-X.
Which machine learning libraries are supported?
It supports algorithms through integrations with Stable Baselines 3, RLlib, and ElegantRL.
