---
id: flaml
version: "2.6.0"
license: MIT
license_treatment: permissive
maintenance: active
---

# FLAML — A fast library for automated machine learning and tuning

License: permissive · Maintenance: active · Downloads: 649.1K/mo

## What it is and what it does

FLAML is a Python library that automates machine learning workflows by handling model selection and hyperparameter optimization. It works as a scikit-learn-style estimator for standard classification and regression tasks, but also supports generic hyperparameter tuning for custom functions and existing ML libraries like XGBoost and LightGBM. The library is designed to work efficiently under resource constraints, making it practical for scenarios where computational budget is limited.

The package depends only on NumPy at runtime and supports Python 3.10 through 3.13. It offers multiple usage patterns: direct AutoML for quick model discovery, restricted learner lists for tuning specific algorithms, custom function tuning, and zero-shot AutoML that applies pre-optimized hyperparameter configurations. The library is actively maintained by Microsoft Research with contributions from academic collaborators.

Use it for:

- Quickly find a well-tuned classification or regression model for a new dataset without manual hyperparameter experimentation.
- Tune hyperparameters for a specific algorithm like LightGBM or XGBoost within a fixed time or resource budget.
- Optimize custom evaluation functions or ML pipelines where standard AutoML tools don't apply.
- Apply pre-optimized hyperparameter configurations to new datasets using zero-shot AutoML patterns.
- Automate hyperparameter tuning for foundation models and LLM inference configurations in production workflows.

## Worth the install?

AI-flagged interpretation of the facts above — verify before relying.

FLAML automates machine learning workflows by selecting models and optimizing hyperparameters under resource constraints, supporting both standard ML tasks and custom function tuning.

Yes. FLAML is actively maintained, has low install friction, carries a permissive MIT license, and offers practical automation for both standard ML tasks and custom tuning scenarios. The single NumPy dependency and support for current Python versions make it straightforward to integrate. No known vulnerabilities and strong community adoption (4385 stars) support its reliability.

## Install

pip install flaml
uv add flaml
poetry add flaml

## Installing FLAML

Before you install: Low friction install with only NumPy as a runtime dependency. Active maintenance with recent release (108 days ago) and 4385 repository stars indicate ongoing development and community adoption.

License in practice: MIT license permits commercial and private use with minimal restrictions, making it suitable for most production and research contexts.

Quickstart:

pip install flaml
from flaml import AutoML
automl = AutoML()
automl.fit(X_train, y_train, task="classification")

Requires Python >= 3.10 and < 3.14; full model support not guaranteed on other versions.

Verify before relying:

- Whether the 'extra options' installation variants (e.g., flaml[automl]) introduce additional dependencies beyond NumPy.
- Performance characteristics and typical runtime for hyperparameter tuning on datasets of varying sizes.
- Specific ML frameworks and estimators supported by default versus those requiring optional dependencies.

## Package facts

- License: MIT (permissive)
- Python support: supports_current
- Install friction: low
- Maintenance: active
- Downloads: 649.1K/month (top 15,000 on PyPI)
- Known vulnerabilities: none known

## Tags

automated machine learning automl, hyperparameter tuning optimization, model selection and tuning, fast automl library, resource-constrained ml automation, hyperparameter optimization framework, ml workflow automation, automl, hyperparameter-tuning, model-selection

[View on SkillFed](https://skillfed.io/packages/flaml) · [View on PyPI](https://pypi.org/project/flaml/)