PyCaret

Spec sheet

What it does

PyCaret is a low-code Python AutoML library and self-hosted machine-learning platform.Read 15 Oct 2026 · github.com

Purpose

PyCaret is an open-source, low-code Python tool for automating machine-learning workflows.Read 7 Oct 2026 · pycaret.org

Deployment

Models can be saved as portable scikit-learn pipelines, served behind FastAPI, or deployed through the optional pycaret-server layer.Read 15 Oct 2026 · pycaret.org

Integrations

PyCaret wraps or interoperates with scikit-learn, XGBoost, Microsoft LightGBM, spaCy, sktime, pyod, FastAPI, React, and Plotly.Read 15 Oct 2026 · linkedin.com

Security

The overview says the self-hosted backend can be wired to existing authentication and is JWT-friendly.Read 7 Oct 2026 · pycaret.org

Support

The FAQ directs users to GitHub Discussions for questions and GitHub Issues for bug reports.Read 7 Oct 2026 · pycaret.org

Headquarters

Toronto, CanadaRead 28 Sep 2026 · pycaret.org

Founded

LinkedIn lists PyCaret as founded in 2020.Read 15 Oct 2026 · linkedin.com

Maker

The FAQ’s citation entry names Moez Ali and contributors as authors.Read 7 Oct 2026 · pycaret.org

Anomaly detection

AnomalyExperiment uses pyod adapters including Isolation Forest, KNN, COPOD, ECOD, LOF, PCA, MCD, ABOD, CBLOF, HBOS, OCSVM, SOS, and SOD.Read 15 Oct 2026 · pycaret.org

Backend API

The optional pycaret-server backend is a FastAPI service with CRUD, run orchestration, WebSockets, pipeline deployment, and plot endpoints.Read 15 Oct 2026 · pycaret.org

Classification

ClassificationExperiment handles categorical targets and includes metrics such as accuracy, AUC, recall, precision, F1, kappa, and MCC.Read 15 Oct 2026 · pycaret.org

Clustering

ClusteringExperiment offers KMeans, affinity propagation, mean shift, spectral, hierarchical, DBSCAN, OPTICS, Birch, KMode, and HDBSCAN adapters.Read 15 Oct 2026 · pycaret.org

Company description

LinkedIn describes PyCaret as an open-source, low-code machine-learning library and a nonprofit software-development organization.Read 15 Oct 2026 · linkedin.com

Company location

LinkedIn lists Toronto, Canada as PyCaret's headquarters and primary location.Read 15 Oct 2026 · linkedin.com

Company size

LinkedIn lists PyCaret's company size as 11–50 employees.Read 15 Oct 2026 · linkedin.com

Compatibility break

Version 4.0 removes the 3.x module-level functional API and requires the new Experiment object-oriented API; mixing 3.x and 4.0 is unsupported.Read 15 Oct 2026 · pycaret.org

Compatibility limit

PyCaret 4.0 removes the 3.x module-level functional API, so the FAQ advises pinning 3.x for existing code and using 4.0 for new work.Read 7 Oct 2026 · pycaret.org

Current maturity

The GitHub repository labels the 4.0 line as work in progress and says the 3.x line is frozen at 3.4.0.Read 15 Oct 2026 · github.com

Dashboard

An optional React dashboard and FastAPI backend can be self-hosted for multi-user and deployment workflows.Read 7 Oct 2026 · pycaret.org

Documentation depth

The official site includes concepts, tutorials, preprocessing references, function guides, FAQs, a cheat sheet, and generated API reference.Read 15 Oct 2026 · pycaret.org

Explainability and plots

The plotting modules expose diagnostic charts for classification, regression, features, clustering, anomalies, time series, and exploratory analysis.Read 15 Oct 2026 · pycaret.org

Forecasting

TimeSeriesExperiment is built on sktime and supports models such as ARIMA, AutoARIMA, exponential smoothing, ETS, Theta, STLF, Croston, and reduced-regressor adapters.Read 15 Oct 2026 · pycaret.org

Hardware

The default path runs on CPU, and GPU-accelerated estimators are available when the relevant libraries are installed.Read 7 Oct 2026 · pycaret.org

Hyperparameter tuning

tune_model performs configurable iterative hyperparameter searches and returns tuned pipelines and parameters.Read 15 Oct 2026 · pycaret.org

Installation

The documented installation is pip install pycaret, with dashboard, explainability, and forecasting extras available.Read 15 Oct 2026 · pycaret.org

Known limitations

The 4.0 changelog lists incomplete plot rewrites, legacy internals still being replaced, and no Python 3.14 support.Read 15 Oct 2026 · pycaret.org

Licence

PyCaret 4.0 licenses the engine under MIT and the control-plane components under BUSL-1.1 with a later Apache-2.0 change date and an additional-use grant.Read 15 Oct 2026 · github.com

Limit: comparative performance

A 2023 healthcare study reported PyCaret accuracy below AutoGluon on several Cleveland, combined, and Hungarian dataset tests, with Hungarian unreduced accuracy around 68%.Read 15 Oct 2026 · scribd.com

Limit: dependency installation

A Stack Overflow question describes repeated setup failures caused by incompatible scikit-learn, pandas-profiling, MarkupSafe, NumPy, and PyCaret versions.Read 15 Oct 2026 · stackoverflow.com

Limit: runtime stability

A Reddit user reported intermittent unknown exceptions during forecasting predictions that were difficult to reproduce locally.Read 15 Oct 2026 · reddit.com

ML tasks

It provides experiment APIs for classification, regression, clustering, anomaly detection, and time-series forecasting.Read 7 Oct 2026 · pycaret.org

Model comparison

compare_models trains registered candidates and returns ranked leaderboards plus fitted top models.Read 15 Oct 2026 · pycaret.org

Notable release

The changelog records PyCaret 4.0.0a6 on May 8, 2026, restoring several legacy verbs and adding a SHAP interpretation extra.Read 15 Oct 2026 · pycaret.org

Package adoption

PePy reports more than 23.5 million PyPI downloads and identifies version 4.0.0a8 as released on May 8, 2026.Read 15 Oct 2026 · pepy.tech

Pipelines

Fitted outputs are portable scikit-learn pipelines, or forecasting pipelines for time-series work.Read 7 Oct 2026 · pycaret.org

Praise: Docker usability

A Reddit commenter said the standard PyCaret Docker image provided a better user experience and less breakage than dependency-heavy local setups.Read 15 Oct 2026 · reddit.com

Preprocessing

The default fit pipeline performs train/test splitting, numeric mean imputation, categorical mode imputation, ordinal encoding, and classification target encoding.Read 15 Oct 2026 · pycaret.org

Public API

The typed pycaret.api package exposes JSON-serializable model, metric, and setup-parameter introspection for UIs and LLM agents.Read 15 Oct 2026 · pycaret.org

Python support

The installation page lists Python 3.11, 3.12, and 3.13 as supported.Read 7 Oct 2026 · pycaret.org

Rating on G2

0.0 out of 5 from 0 ratings; the G2 category page provides no user-review information for PyCaret.Read 15 Oct 2026 · g2.com

Regression

RegressionExperiment handles continuous targets and provides linear, tree, boosting, neural-network, and kernel models.Read 15 Oct 2026 · pycaret.org

Security controls

The repository documents secret scanning, Fernet encryption for stored LLM keys and passwords, JWT signing, and exclusion of database and credential files from commits.Read 15 Oct 2026 · github.com

Self-hosting

The repository provides a Docker Compose installation with React web UI, FastAPI backend, SQLite data volume, and local model artifacts.Read 15 Oct 2026 · github.com

Support channels

The FAQ directs users to GitHub Discussions for questions, GitHub Issues for bugs, and LinkedIn for announcements.Read 15 Oct 2026 · pycaret.org

System requirements

PyCaret 4.0 supports Python 3.11, 3.12, and 3.13; CPU is sufficient, GPU acceleration is optional, and 4 GB memory is described as comfortable for tutorial data.Read 15 Oct 2026 · pycaret.org

Task coverage

The engine supports classification, regression, clustering, anomaly detection, and time-series forecasting.Read 15 Oct 2026 · pycaret.org

Workflow

The engine supports fitting experiments, comparing and tuning models, making predictions, and saving or loading models.Read 7 Oct 2026 · pycaret.org

Plans and prices

Every PyCaret plan and price →

Where Specifiction places it

• Predictive Analytics Software#5 of 378.4