xgrammar 0.2.8

Efficient, Flexible and Portable Structured Generation

Get Started | Documentation | Blogpost | Technical Report

News

• [2026/5] XGrammar-2 has been released! Check out our blog for more information.
• [2025/12] XGrammar has been officially integrated into Mirai
• [2025/09] XGrammar has been officially integrated into OpenVINO GenAI
• [2025/02] XGrammar has been officially integrated into Modular's MAX
• [2025/01] XGrammar has been officially integrated into TensorRT-LLM.
• [2024/12] XGrammar has been officially integrated into vLLM.
• [2024/12] We presented research talks on XGrammar at CMU, UC Berkeley, MIT, THU, SJTU, Ant Group, LMSys, Qingke AI, Camel AI. The slides can be found here.
• [2024/11] XGrammar has been officially integrated into SGLang.
• [2024/11] XGrammar has been officially integrated into MLC-LLM.
• [2024/11] We officially released XGrammar v0.1.0!

Overview

XGrammar is an open-source library for efficient, flexible, and portable structured generation.

It leverages constrained decoding to ensure 100% structural correctness of the output. It supports general context-free grammar to enable a broad range of structures, including JSON, regex, custom context-free grammar, etc.

XGrammar uses careful optimizations to achieve extremely low overhead in structured generation. It has achieved near-zero overhead in JSON generation, making it one of the fastest structured generation engines available.

XGrammar features universal deployment. It supports:

• Platforms: Linux, macOS, Windows
• Hardware: CPU, NVIDIA GPU, AMD GPU, Apple Silicon, TPU, etc.
• Languages: Python, C++, JavaScript, and Swift APIs
• Models: Qwen, Llama, DeepSeek, Phi, Gemma, etc.

XGrammar is very easy to integrate with LLM inference engines. It is the default structured generation backend for most LLM inference engines, including vLLM, SGLang, TensorRT-LLM, and MLC-LLM, as well as many other companies. You can also try out their structured generation modes!

Get Started

Install XGrammar:

pip install xgrammar

For use with MPS on Apple Silicon, install with:

pip install "xgrammar[metal]"

Import XGrammar:

import xgrammar as xgr

Please visit our documentation to get started with XGrammar.

• Installation
• Quick start

Contributing and governance

We welcome contributions of code, tests, documentation, bug reports, and reviews. See CONTRIBUTING.md to get started. All participants are expected to follow our Code of Conduct.

GOVERNANCE.md describes the current maintainers, technical decision process, and maintainer role changes. CODEOWNERS lists the reviewers responsible for specific parts of the repository.

Releases

XGrammar publishes versioned releases on GitHub Releases and Python packages on PyPI. Releases are made as needed when features, bug fixes, or compatibility updates are ready, without a fixed calendar schedule. Prereleases are identified separately from stable releases.

For each release, a maintainer selects a reviewed commit from main and coordinates the following:

1. Verify that the applicable automated tests and package builds pass for the release commit, and resolve failures that affect the release.
2. Prepare release notes describing features, fixes, compatibility changes, and known limitations.
3. Tag the release commit and publish a GitHub release. The package workflow builds wheels and a source distribution and publishes them to PyPI when the GitHub release is published. Maintainers can also invoke that workflow manually.
4. Check that the published packages and release notes are available, and follow up on reported regressions with fixes or a subsequent release.

The Python package version is generated from Git tags by setuptools-scm; no manual version update in pyproject.toml is needed. For example, v0.2.7 builds as 0.2.7, and commits after a release get development versions. Builds from Git need full history and tags; published source distributions retain their version without Git. To publish manually, run the package workflow on a release tag; manual branch runs only build artifacts.

Third-Party Bindings

• Rust: xgrammar-rs — Community Rust bindings for XGrammar.

Collaborators

XGrammar has been widely adopted in industry, open-source projects, and academia. Our collaborators include:

Citation

If you find XGrammar useful in your research, please consider citing our papers:

@article{dong2024xgrammar,
  title={Xgrammar: Flexible and efficient structured generation engine for large language models},
  author={Dong, Yixin and Ruan, Charlie F and Cai, Yaxing and Lai, Ruihang and Xu, Ziyi and Zhao, Yilong and Chen, Tianqi},
  journal={Proceedings of Machine Learning and Systems 7},
  year={2024}
}
@inproceedings{10.1145/3786335.3813124,
  author = {Li, Linzhang and Dong, Yixin and Wang, Guanjie and Xu, Ziyi and Jiang, Alexander and Chen, Tianqi},
  title = {XGrammar-2: Dynamic and Efficient Structured Generation Engine for Agentic LLMs},
  year = {2026},
  isbn = {9798400724152},
  publisher = {Association for Computing Machinery},
  address = {New York, NY, USA},
  url = {https://doi.org/10.1145/3786335.3813124},
  booktitle = {Proceedings of the ACM Conference on AI and Agentic Systems},
  pages = {1009--1022},
  numpages = {14}
}