---
id: conllu
version: "6.0.0"
license: The MIT License (MIT) Copyright (c) 2016 Emil Stenström Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the… (full text in the JSON record)
license_treatment: unclear
maintenance: active
---

# conllu — CoNLL-U Parser parses a CoNLL-U formatted string into a nested python dictionary

License: unclear · Maintenance: active · Downloads: 405.9K/mo

## What it is and what it does

conllu is a lightweight parser that converts CoNLL-U formatted text—a standard annotation format used in natural language processing and linguistic research—into Python data structures. CoNLL-U is the output format of many NLP pipelines and dependency parsers, encoding tokens with their linguistic properties (part-of-speech tags, lemmas, morphological features, syntactic dependencies). The parser converts this text into nested dictionaries and TokenList objects that you can query, filter, and manipulate programmatically.

The package has zero runtime dependencies and is designed for simplicity: it's about 300 lines of code with full type hints for IDE autocompletion. It supports both one-shot parsing of entire files and incremental parsing via generators for memory efficiency on large datasets. You can filter tokens by field values (including nested features), modify parsed structures, and serialize them back to CoNLL-U format.

Use it for:

- Parse output from dependency parsers (like spaCy or UDPipe) to extract and analyze syntactic relationships in text.
- Filter linguistic annotations by part-of-speech, morphological features, or dependency relations for corpus analysis.
- Modify or augment CoNLL-U annotations programmatically and write the results back to file.
- Process large annotated corpora incrementally without loading entire files into memory.
- Build NLP pipelines that consume Universal Dependencies formatted data as intermediate representations.

## Worth the install?

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

Parses CoNLL-U formatted text (a standard NLP annotation format) into nested Python dictionaries with full type support, enabling programmatic access to linguistic dependency structures.

Yes. Zero dependencies, active maintenance, no known vulnerabilities, and MIT licensing make it a safe, lightweight choice. Install it if you work with CoNLL-U data from NLP tools or linguistic corpora. The only gotcha is the Python 3.8+ requirement; if you're on an older version, pin to conllu==4.5.3 or earlier.

## Install

pip install conllu
uv add conllu
poetry add conllu

## Installing conllu

Before you install: Low friction: no runtime dependencies, pure Python wheel, actively maintained with last commit 2026-08-05. Requires Python 3.8 or later.

License in practice: MIT license permits unrestricted use, modification, and distribution with attribution; no restrictions on commercial or proprietary use.

Quickstart:

from conllu import parse

data = """# text = Example
1\tThe\tthe\tDET\tDT\t_\t2\tdet\t_\t_
2\tdog\tdog\tNOUN\tNN\t_\t0\troot\t_\t_
"""

sentences = parse(data)
print(sentences[0][0]['form'])  # 'The'

Requires Python 3.8 or later; CoNLL-U input must follow the Universal Dependencies format specification.

Verify before relying:

- Whether the package handles malformed or incomplete CoNLL-U input gracefully or raises exceptions.
- Performance characteristics for very large files beyond the mentioned 'over a megabyte' threshold.
- Whether full type support includes runtime type checking or only static type hints for editors.

## Package facts

- License: The MIT License (MIT) Copyright (c) 2016 Emil Stenström Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the… (full text in the JSON record) (unclear)
- Python support: supports_current
- Install friction: low
- Maintenance: active
- Downloads: 405.9K/month (top 15,000 on PyPI)
- Known vulnerabilities: none known

## Tags

conll-u parser, parse conll-u format, nlp dependency parsing, linguistic annotation parser, universal dependencies parser, conll format reader, nlp data structure, nlp, parsing, linguistics

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