---
id: autogluon-text
version: "0.6.2"
license: Apache-2.0
license_treatment: permissive
maintenance: active
---

# autogluon.text — AutoML for Image, Text, and Tabular Data

License: permissive · Maintenance: active · Downloads: 84.6K/mo

## What it is and what it does

autogluon.text is a specialized AutoML package for text classification and prediction tasks. It wraps AutoGluon's text prediction capabilities, allowing you to train deep learning models on text data with just a few lines of code—no manual feature engineering or hyperparameter tuning required. The package handles model selection, training, and evaluation automatically, targeting use cases where you have labeled text data and want a production-ready predictor quickly.

The package depends on autogluon.multimodal, which provides the underlying neural network infrastructure. It's designed for developers and data scientists who want to avoid the complexity of building text models from scratch but need reasonable accuracy without extensive ML expertise. The API mirrors AutoGluon's other predictors (TabularPredictor, ImagePredictor), making it familiar if you've used other AutoGluon modules.

Use it for:

- Train a text classifier on customer feedback or support tickets to automatically categorize incoming messages.
- Build a sentiment analysis model on labeled reviews without writing custom neural network code.
- Quickly prototype a text-based prediction model for a Kaggle competition or proof-of-concept.
- Deploy a production text classifier that automatically selects and tunes models for your specific dataset.
- Combine text predictions with other data types using AutoGluon's multimodal capabilities.

## Worth the install?

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

Automates machine learning for text data, training and deploying high-accuracy models with minimal code through AutoGluon's text prediction interface.

Yes, if you need to train text models quickly without deep ML expertise and can work within Python 3.7–3.9. The low install friction, active maintenance, permissive license, and zero known vulnerabilities make it a solid choice for prototyping and production use. No, if you require the latest Python versions or need fine-grained control over model architecture—use lower-level libraries like transformers or PyTorch instead.

## Install

pip install autogluon-text
uv add autogluon-text
poetry add autogluon-text

## Installing autogluon.text

Before you install: Low install friction with a single runtime dependency (autogluon.multimodal). The package is actively maintained with recent commits and has been in active development since its first release in 2020.

License in practice: Licensed under Apache-2.0 (permissive), allowing free use, modification, and distribution with minimal restrictions—suitable for both open-source and commercial projects.

Quickstart:

pip install autogluon.text

from autogluon.text import TextPredictor
predictor = TextPredictor(label='class').fit(train_data, time_limit=120)
leaderboard = predictor.leaderboard(test_data)

Requires Python 3.7–3.9 (capped below current versions); check your environment before installing.

Verify before relying:

- Whether autogluon.multimodal's own dependencies introduce significant additional install friction or system requirements.
- Current state of model download sizes and memory requirements for typical text prediction tasks.
- Performance characteristics and training time for different text dataset sizes.

## Package facts

- License: Apache-2.0 (permissive)
- Python support: capped_below_current
- Install friction: low
- Maintenance: active
- Downloads: 84.6K/month (top 15,000 on PyPI)
- Known vulnerabilities: none known

## Tags

automl text classification, automated text prediction, text nlp machine learning, deep learning text models, text data automl, neural text classification, automated nlp training, automl, text-classification, nlp

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