facebookresearch/fastText

Health v2 · maintenance only

Flags: archived

Adoption not part of the score

26534 stars · 4837 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

fastText is a lightweight open-source C++ library (with Python bindings and a CLI) from Facebook Research for efficiently learning word representations and training text classifiers. It ships pre-trained word vectors for 157 languages and supports model quantization so models can run on modest hardware or mobile devices.

Use cases

• train word embeddings for my corpus
• classify documents into categories like spam vs not spam
• get word vectors for out-of-vocabulary or misspelled words
• detect the language of a text snippet
• run sentiment analysis on product reviews
• compress a text classification model to fit on mobile
• download pre-trained word vectors for 157 languages

When to choose

• you need fast, lightweight text classification or word vectors on CPU-only hardware
• you need embeddings for rare or misspelled words via subword (character n-gram) information
• you want pre-trained vectors for many languages or a small language-identification model
• you need models small enough for mobile or embedded deployment via quantization

When to avoid

• you need contextual/transformer-based embeddings for state-of-the-art NLU
• you need deep learning pipelines, GPU training, or fine-tuning of large models
• you primarily develop on Windows, which is not officially supported
• you need actively developed features - the project is largely in maintenance mode

Facets

library · maturity maintenance

nlp machine-learning cli machine-learning developer-tools cpp python cli word-embeddings word-vectors text-classification sentence-classification subword-ngrams pretrained-models language-identification model-quantization natural-language-processing linux macos

Member repositories

For agents

markdown · JSON · MCP: product_card(name="facebookresearch/fastText")