Who uses psutil

psutil is among the top 100 most-downloaded packages on PyPI, with 280+ million downloads per month and 790,000+ GitHub repositories using it. The projects below are a sample of notable software that uses it. They are grouped by how much of the project depends on psutil, and every usage links to the code that calls psutil. See also Alternatives for related Python libraries and equivalents in other languages.

Agent checks read processes, disks and network through psutil.
Ray
Ships psutil in its own wheel; OOM monitor, process watcher, autoscaler.
CPU, memory and disk for the system monitor integration.
Its system-metrics instrumentation, which is how Splunk and Honeycomb users get psutil.
Every metric it displays comes from psutil.

Built on psutil

psutil is a core dependency of these projects and central to how they work.

Project

Description

What it uses psutil for

Ray

Distributed runtime for AI and Python workloads

bundled inside the wheel; OOM monitor, process watcher, autoscaler, dashboard

Glances

System monitoring tool (top/htop alternative)

CPU, memory, disk, network, sensors, processes

Salt

Infrastructure automation at scale

master, minion, the ps execution module, beacons, grains

bpytop

Terminal resource monitor

CPU, memory, disk, network, processes

auto-cpufreq

CPU speed and power optimizer for Linux

CPU load and frequency readings

s-tui

Terminal CPU stress and monitoring utility

CPU, frequency and temperature readings

GRR

Remote live forensics by Google

client actions for processes, network connections and memory

psdash

Web dashboard using psutil and Flask

CPU, memory, disk, network, users, processes

OpenTelemetry Python

Observability instrumentation for Python

the whole instrumentation-system-metrics package; Splunk, Honeycomb and New Relic users depend on psutil through it

psleak

Test framework to detect memory leaks in Python C extensions

heap process memory (heap_info())

Used for a specific feature

In these, psutil backs a single feature rather than the whole project.

Project

Description

What it uses psutil for

Home Assistant

Home automation platform

the system monitor integration

Ansible

IT automation platform

counting active connections in the wait_for module

Apache Airflow

Workflow orchestration platform

task subprocess lifecycle

Sentry

Error tracking and performance monitoring

telemetry metrics

Celery

Distributed task queue

memory sampling used to debug leaks

MLflow

Machine learning lifecycle platform

system metrics logged with each run

Locust

Scalable load testing in Python

monitoring its own process

Dask

Parallel computing with task scheduling

worker memory management in distributed

Accelerate

Distributed training and inference launcher

physical CPU count, available memory, process launching

Spyder

Scientific Python IDE

memory usage in the status bar

Ajenti

Web-based server administration panel

dashboard widgets for CPU, memory, uptime and traffic

asitop

Apple Silicon performance monitoring CLI

memory and swap readings

integrations-core

Integration checks for the Datadog agent

15 check modules, including process, disk and network

Elastic APM

Application performance monitoring agent

CPU and memory metrics on Windows and macOS (its own /proc parser on Linux)

Used in development only

psutil is a test dependency only. Users of these projects never install it.

Project

Description

What it uses psutil for

TensorFlow

Machine learning framework by Google

unit tests

PyTorch

Deep learning framework with GPU acceleration

benchmark scripts

How this list was compiled

Candidates came from the GitHub dependency graph and from code search. Each one was then checked against the project’s packaging metadata and source, to make sure psutil is a real dependency and not just a mention. The links in the last column point at the code doing the work, pinned to a commit so the line numbers keep matching.