pynini
Finite-state grammar compilation
Decision gist · record as of 2026-08-14
Yes, if you are working on computational linguistics, morphology, or formal language tasks that map naturally to finite-state automata. The library is stable, permissively licensed, and well-suited for research and production use. No, if your project does not require explicit FST manipulation or if you lack the ability to install OpenFst and a C++17 compiler on your system—the installation friction is non-trivial for casual users.AI-flagged interpretation of the facts on this page — verify before relying
Before you install
- OpenFst 1.8.4 with grm extensions must be installed and discoverable by the build system; a C++17 compiler (GCC ≥7 or Clang ≥700) is required to build from source.
- Medium friction: requires a C++17 compiler and OpenFst 1.8.4 with grm extensions pre-installed on the system.
- Precompiled manylinux wheels available for Python 3.9–3.14 on x86_64 Linux, but Windows and macOS users should use conda-forge.
License · maintenance · safety
Apache 2.0 (permissive) — Apache 2.0 permissive license allows commercial and private use with minimal restrictions. You must retain license notices but face no copyleft obligations.
last release 2025-09-04 (344 days)
0 known vulnerabilities (OSV.dev, 2026-08-14) · 239,094 downloads/mo, #8,925 on PyPI
Alternatives
Verify before relying
pip install pynini
import pynini
# Create a simple FST
fst = pynini.string_map([("hello", "hi")])
result = pynini.shortestpath(fst)- Whether precompiled wheels include OpenFst or if it must be installed separately on Linux.
- Current maintenance status and whether the 344-day gap since last release signals active development or dormancy.
- Availability of comprehensive documentation or tutorials beyond the cited 2016 paper.
What it is and what it does
Pynini is a Python extension module that compiles linguistic grammar rules into weighted finite-state transducers (FSTs), pushdown transducers, and multi-pushdown transducers. It wraps OpenFst and provides a high-level interface for constructing and manipulating automata-based language models. The library is designed for computational linguists and NLP researchers who need to express complex morphological, phonological, or syntactic transformations as formal automata.
The package has no runtime dependencies beyond OpenFst itself (which must be installed separately on most platforms). Installation on Linux via pip uses precompiled manylinux wheels for Python 3.9–3.14, but Windows and macOS users typically rely on conda-forge. The library is stable and production-ready, though the last release was 344 days ago, suggesting a mature but slowly-evolving codebase.
Use it for
- Build morphological analyzers and generators for inflectional languages by encoding rules as weighted transducers.
- Implement phonological rule systems that transform phonemic representations into surface forms.
- Compile domain-specific grammar rules for text normalization, transliteration, or dialect conversion.
- Prototype and test finite-state models of linguistic phenomena before deploying them in larger NLP pipelines.
- Combine multiple FSTs to create complex multi-stage language processing workflows.
Worth the install?
AI-flagged interpretation of the facts on this page. Verify before relying on it.
Yes, if you are working on computational linguistics, morphology, or formal language tasks that map naturally to finite-state automata.
The library is stable, permissively licensed, and well-suited for research and production use. No, if your project does not require explicit FST manipulation or if you lack the ability to install OpenFst and a C++17 compiler on your system—the installation friction is non-trivial for casual users.
Install
pynini on PyPI
Before you install
Medium friction: requires a C++17 compiler and OpenFst 1.8.4 with grm extensions pre-installed on the system. Precompiled manylinux wheels available for Python 3.9–3.14 on x86_64 Linux, but Windows and macOS users should use conda-forge. Last release 344 days ago; maintenance status is aging.
OpenFst 1.8.4 with grm extensions must be installed and discoverable by the build system; a C++17 compiler (GCC ≥7 or Clang ≥700) is required to build from source.
License in practice
Apache 2.0 permissive license allows commercial and private use with minimal restrictions. You must retain license notices but face no copyleft obligations.
Quickstart
pip install pynini
import pynini
# Create a simple FST
fst = pynini.string_map([("hello", "hi")])
result = pynini.shortestpath(fst)Verify before relying
- Whether precompiled wheels include OpenFst or if it must be installed separately on Linux.
- Current maintenance status and whether the 344-day gap since last release signals active development or dormancy.
- Availability of comprehensive documentation or tutorials beyond the cited 2016 paper.
Package facts
| License | Apache 2.0 permissive |
| Python support | Supports the current Python release >=3.8 |
| Install friction | Medium. Platform-specific wheel |
| Runtime dependencies | None |
| Maintenance | Aging 344 days since the last release |
| First released | |
| Downloads | 239,094 / month, #8,925 on PyPI 30-day window, as of 2026-08-14 |
| Known vulnerabilities | None known OSV.dev, checked 2026-08-14 |
| Classifiers | Development Status :: 5 - Production/StableEnvironment :: ConsoleEnvironment :: Other EnvironmentIntended Audience :: DevelopersLicense :: OSI Approved :: Apache Software LicenseOperating System :: OS IndependentProgramming Language :: Python :: 3.10Programming Language :: Python :: 3.11Programming Language :: Python :: 3.12Programming Language :: Python :: 3.8Programming Language :: Python :: 3.9Topic :: Scientific/Engineering :: Artificial IntelligenceTopic :: Scientific/Engineering :: MathematicsTopic :: Software Development :: Libraries :: Python ModulesTopic :: Text Processing :: Linguistic |
Evidence: pynini-2.1.7-cp310-cp310-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl; pynini-2.1.7-cp311-cp311-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl; pynini-2.1.7-cp312-cp312-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl; pynini-2.1.7-cp313-cp313-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl; pynini-2.1.7-cp313-cp313t-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl; pynini-2.1.7-cp314-cp314-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl; pynini-2.1.7-cp314-cp314t-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl; pynini-2.1.7-cp39-cp39-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl; pynini-2.1.7-pp310-pypy310_pp73-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl; pynini-2.1.7-pp311-pypy311_pp73-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl
Tags
Let your AI agent find packages like this
Example. Real query, live index.
You found this page by searching. An agent finds it by wishing: SkillFed indexes 14,416 PyPI packages by what they can do, searchable in plain language.
wish › “finite-state transducers python”
- pyniniPynini compiles grammar rules into weighted finite-state transducers…
- kaldifstPython wrapper for OpenFst and Kaldi extensions, providing…
- nemo-text-processingConverts text to normalized form and back (inverse normalization) for…
Give your agent the search over MCP, or paste the wish link into any chat.
More Python Modules packages
Converts domain names between Unicode and ASCII-compatible encoding (Punycode) according to IDNA 2008 and Unicode Technical Standard 46, with security validation and broader script coverage than the standard library.
Install it if you work with internationalized domain names, need to validate domains, or use HTTP clients that depend on it transitively.
Setuptools is a Python build backend and package management tool that handles building, distributing, and installing Python packages, including support for C/C++ extension modules.
PyYAML parses and emits YAML 1.1 data format, enabling serialization and deserialization of configuration files and Python objects to and from human-readable YAML text.
Pydantic validates Python data structures against type hints, coercing and checking input at runtime to ensure it matches a declared schema.
Provides reusable metadata objects for use with PEP-593 `typing.Annotated` to express common constraints like bounds, collection sizes, and predicates on types.
Install it if you use or build libraries that need to express type constraints in a standardized, inspectable way—or if you want to annotate your own types with…
Provides runtime tools to inspect and introspect Python type annotations, enabling programmatic examination of type hints at execution time.
See also nemo-text-processing · PyMeta3 · Automat · konlpy · wetext · kaldifst · nltk · TatSu · Parsley · tree-sitter-python