CVE-2025-58446 xgrammar vulnerable to denial of service by huge enum grammar
MediumCVSS 6.9
EPSS 0.62%
- Published
- Updated
- Exploit status
- No public exploit known
- CISA SSVC
- Proof of concept public · automatable: yes · technical impact: partial
Description
xgrammar is an open-source library for efficient, flexible, and portable structured generation. A grammar optimizer introduced in 0.1.23 processes large grammars (>100k characters) at very low rates, and can be used for DOS of model providers. This issue is fixed in version 0.1.24.
Upgrade path
2 packages: 1 fixed.
| Package | Affected | Fixed in | Status |
|---|---|---|---|
| pypi xgrammar(PyPI) | = 0.1.23; >= 0.1.23, < 0.1.24 | 0.1.24 | Fixed |
| mlc-ai xgrammar | = 0.1.23; = = 0.1.23, < 0.1.24 | No fix listed | Versions listed |
Affected products
| Product | Affected versions | Fixed in |
|---|---|---|
| pypi xgrammar(PyPI) | >= 0.1.23, < 0.1.24 | 0.1.24 |
| pypi xgrammar(PyPI) | = 0.1.23 | No fix listed |
| mlc-ai xgrammar | = = 0.1.23, < 0.1.24 | No fix listed |
| mlc-ai xgrammar | = 0.1.23 | No fix listed |
References
- https://github.com/mlc-ai/xgrammar/security/advisories/GHSA-9q5r-wfvf-rr7f exploit vendor-advisory
- nvd.nist.gov third-party-advisory vendor-advisory
- https://github.com/mlc-ai/xgrammar/commit/ced69c3ad2f8f61b516cc278a342e7c644383e27 patch
- github.com product
- pypi.org product
- github.com vendor-advisory