CVE-2026-25048

xgrammar: Multi-layer nesting causes DoS

HIGH

7.5

CVSS Score

Published: March 5, 2026 at 03:34 PM UTC

Updated: July 21, 2026 at 01:39 PM UTC

3 products affected across 2 vendorsXgrammarEnterprise Linux AIAI Inference Server

Using xgrammar by mlc-ai?

Get instant alerts when new vulnerabilities affect xgrammar - before attackers find them.

Real-time alertsTrack your software stackFree plan available

Start Free Sign In

Severity

HIGH7.5

Exploit prediction

EPSS 1%

SSVC

No known exploit

Description

xgrammar is an open-source library for efficient, flexible, and portable structured generation. Prior to version 0.1.32, the multi-level nested syntax caused a segmentation fault (core dumped). This issue has been patched in version 0.1.32.

Affected

Our own answer: ranges merged, affected separated from platform-only, with confidence and review flags.

Xgrammar · Mlc AI

95%

v < 0.1.32

Affected range stops below 0.1.32, so it looks safe — our inference, not a vendor statement

AI Inference Server · Red Hatcontainer

60%

All versions

Vendor states this is fixed in 1774351144 (unaffected)

Vendor states this is fixed in 1774547384 (unaffected)

Vendor states this is fixed in 1775252598 (unaffected)

Enterprise Linux AI · Red Hatrpm

60%

Remediation

No guide yet

No remediation guidance has been written for this CVE yet. What the record does show:

Patch and advisory references

https://github.com/mlc-ai/xgrammar/security/advisories/GHSA-7rgv-gqhr-fxg3

References (10)

x_refsource_CONFIRMExploitVendor Advisory

Source: cverecord

https://github.com/mlc-ai/xgrammar/security/advisories/GHSA-7rgv-gqhr-fxg3

x_refsource_MISCRelease Notes

Source: cverecord

https://github.com/mlc-ai/xgrammar/releases/tag/v0.1.32