PM4Py
processintelligence.solutions
- Free plan
- Windows
- macOS
- Linux
- API
What PM4Py is
PM4Py is a Python library that lets developers run process mining algorithms from their own code. It can derive directly-follows graphs, Petri nets, BPMN models, and process trees from event data, then compare observed activity with process models through replay and alignment-based measures. Its object-centric mining can handle logs in which events are linked to multiple object types, such as orders, items, and deliveries. PM4Py imports and exports event logs and supports XES and OCEL 1.0 and 2.0. It also extracts machine-learning features for use with pandas and scikit-learn. PM4Py is free under the Community Version, licensed as AGPL v3. Covered work must be published under AGPL when distributed or made available over a network; a commercial license allows closed-source use. The commercial option includes regular releases, early feature access, and dedicated support, with pricing available by quote. It runs on Windows, Linux, and macOS; pip is recommended. Graphviz is needed for visualization.
Who it suits
PM4Py suits developers, data scientists, and researchers who want process mining within Python workflows. Teams planning closed-source commercial use should consider the commercial license and its quote-based pricing.
Good to know, and what to watch
- Discovers graphs, Petri nets, BPMN models, and process trees.
- Supports replay and alignment-based conformance measures.
- Handles multi-object event logs with XES and OCEL support.
- Extracts features for pandas and scikit-learn workflows.
- Available free under the Community Version.
- AGPL-covered distributed or network-accessible work must use AGPL.
- Graphviz is required for visualization features.
- Community version lists no support.
- Commercial license pricing is by quote.
Plans
Every plan and priceCommercial Version
Community Version
Free plan
Getting started
- Open the PM4Py website
- Install Python dependencies with pip
- Use the Windows or Linux installation instructions
- Install and configure Graphviz for visualization
- Choose the AGPL Community Version or request a Commercial Version quote
Limits to check
The Community Version is free under AGPL v3, and covered software distributed or made available over a network must also be published under AGPL. It does not include listed support; closed-source commercial use requires the Commercial Version or an evaluation license.
The Club's review
Overview
PM4Py is a Python library for building process-mining analysis into code, aimed at developers, data scientists, and researchers. It stands out for its range of process models and its support for logs involving several object types. The main trade-offs are manual data refresh, a Graphviz requirement for visualization, and licensing terms that can complicate closed-source work.
Key features
Discover and check process models
PM4Py can derive directly-follows graphs, Petri nets, BPMN models, and process trees from event data. That variety gives technical users options for examining and representing a process within their own Python workflows. Replay and alignment-based measures compare observed activity with a process model, and root-cause analysis is also supported. Data refresh is manual, so teams planning recurring analysis need to arrange how new event data reaches their workflow.
Handle event logs and connected objects
The library imports and exports event logs, supports XES, and works with OCEL 1.0 and OCEL 2.0. Its object-centric mining can analyze events tied to multiple object types, such as orders, items, and deliveries. This is useful when a single-case view would miss relationships in the event data.
Fit process analysis into Python
PM4Py supports machine-learning feature extraction for use with pandas and scikit-learn, making it a plausible choice for teams combining process analysis with data-science work. Installation is available through pip, with instructions for Windows and Linux. Visualization requires Graphviz to be installed and configured, adding a setup step for users who need diagrams.
Pricing
The Community Version is free of charge at no cost. It is intended for teaching, research, and open-source projects under AGPL v3, with covered work required to be published under AGPL. Community users get no support, so this plan suits projects that can handle setup and maintenance themselves.
The Commercial Version has custom pricing by quote. It permits closed-source commercial use and includes regular releases, early feature access, and dedicated professional support. Commercial evaluation without opening the code requires a specific evaluation license, which can be requested from the maker.
The AGPL terms are a consequential constraint: software using PM4Py must be published under AGPL when distributed or made available over a network. Organizations that need to keep such work closed source should consider the commercial license before building around the community version.
Platforms
PM4Py is available as an API for Linux, macOS, and Windows. It is self-hosted, putting installation and the workflow around manual data refresh in the user's hands. ERP connectors are supported, though the library's Python focus means it is best suited to teams able to work in code.
Who it's for
PM4Py fits developers, data scientists, and researchers who want to call process-mining algorithms from Python, especially when they need several model-discovery options or object-centric analysis. It is less suitable for teams looking for a ready-made interface, automatic data refresh, or community support. Proprietary commercial projects should account for the separate commercial license.
Pros and cons
- Broad model discovery: generates directly-follows graphs, Petri nets, BPMN models, and process trees from event data.
- Object-centric support: handles OCEL 1.0 and 2.0 and events related to multiple object types.
- Python data-science fit: feature extraction can feed pandas and scikit-learn workflows.
- AGPL obligation: distributed or network-accessible covered work must be published under AGPL unless using the commercial license.
- Extra visualization setup: Graphviz must be installed and configured.
- Manual refresh and no community support: recurring updates and troubleshooting fall to the team using the free plan.
Alternatives
Process Mining Software is a useful starting point for readers comparing products across the category.
ProcessMind may suit teams that prefer a web-based option with a free plan: its Free tier allows one member, unlimited models, and 20 AI credits per month, with no extra seats or free viewer seats. It also offers a free trial.
Process.Science is worth considering for teams seeking web, Windows, or self-hosted deployment and predefined report pages, custom visuals, and SQL Server or cloud transformation in its Enterprise plan; pricing is by inquiry.
bupaR is a free and open-source R package for business process analysis, so it is a fit for teams working in R rather than Python.
Celonis Process Mining Platform is another process-mining option with a free plan.
Mindzie Process Mining may suit users who want a desktop edition with full functionality and no time-limited trial; its free edition is capped at 1,000 cases.
Pega CRM is a paid web-based alternative with a free trial.
ProDiscovery offers cloud plans and an on-premise version, with pricing by inquiry.
UiPath Maestro Case Management offers a free trial and a request-demo option.
Verdict
Choose PM4Py if your team is comfortable in Python and needs process discovery, conformance checking, or object-centric event analysis inside its own workflows. Its breadth and data-science integration are strong reasons to use it, but the AGPL obligation is the key reason to look elsewhere or budget for a commercial license; users who need a graphical product or automated refresh should also consider other options.
The verdict
PM4Py combines process discovery, conformance checking, and object-centric analysis in a Python library. Its AGPL terms matter for distributed or network-accessible work, while closed-source commercial use requires a quote-based license.
Questions people ask
How much does PM4Py cost?
The Community Version costs nothing and is free of charge. Commercial pricing is available by requesting a quote.
Which operating systems does PM4Py support?
PM4Py is available for Windows, Linux, and macOS, and it can be used through its API.
What license applies to the free version?
The Community Version uses AGPL v3. Covered work distributed or made available over a network must be published under AGPL.
Can PM4Py be used in closed-source software?
The Commercial Version permits closed-source commercial use. A specific evaluation license is available for commercial evaluation without open-sourcing code.
Does PM4Py support process conformance checking?
Yes. It compares observed behavior with process models using replay and alignment-based measures.
What is needed for PM4Py visualization?
Graphviz must be installed and configured before using PM4Py visualization features.
What the maker publishes
- Audience
- The maker describes PM4Py as built for developers, data scientists, and researchers. source
- Conformance
- It supports comparing observed behavior with process models, including replay and alignment-based measures. source
- Discovery
- It can discover directly-follows graphs, Petri nets, BPMN models, and process trees from event data. source
- Evaluation
- The maker says commercial evaluation without open-sourcing code is available by requesting a specific evaluation license. source
- Event data
- The feature list states that PM4Py imports and exports event logs, supports XES, and supports OCEL 1.0 and OCEL 2.0. source
- Installation
- The maker provides Windows and Linux installation instructions and recommends installing with pip. source
- License limit
- The maker says AGPL-covered software using PM4Py must be published under AGPL when distributed or made available over a network; a commercial license permits closed-source use. source
- Maintenance
- The maker says Process Intelligence Solutions GmbH has been responsible for PM4Py’s development, maintenance, distribution, and licensing since 2024. source
- Object-centric mining
- It can analyze event logs where events relate to multiple object types, such as orders, items, and deliveries. source
- Product
- PM4Py is a Python library that lets developers call process mining algorithms from their own code. source
- Python ecosystem
- The maker shows machine-learning feature extraction for use with pandas and scikit-learn. source
- Support
- The commercial version includes dedicated professional support; the community version lists no support. source
- Visualization requirement
- Graphviz must be installed and configured to use PM4Py’s visualization features. source
Checked against the maker's pages on 7 Oct 2026.