Vulnerability · Published Nov 21, 2025 · Assigned by GitHub_M

CVE-2025-62608MLX has heap-buffer-overflow in load()

Medium 5.5

In plain English

A medium-severity heap buffer overflow in mlx from ml-explore. Proof-of-concept exploit code is public. The chance it is exploited in the next 30 days is estimated at 0.6% (higher than 47% of all known vulnerabilities). Fixed versions are available - upgrade the affected package.

Severity (CVSS 4.0)5.5Medium

Exploit odds (30 days)0.6%top 53.2% of all CVEs

Exploited in the wildNo reportsCISA SSVC: poc

Attack toolingNoneno public tooling known

Official description

MLX is an array framework for machine learning on Apple silicon. Prior to version 0.29.4, there is a heap buffer overflow in mlx::core::load() when parsing malicious NumPy .npy files. Attacker-controlled file causes 13-byte out-of-bounds read, leading to crash or information disclosure. This issue has been patched in version 0.29.4.

Affected packages & fixed versions

Ecosystem	Package	Fixed in	Advisory
PyPI	mlx	0.29.4	GHSA-w6vg-jg77-2qg6
PyPI	mlx	0.29.4	PYSEC-2025-138

Affected products (as reported)

Vendor	Product	Versions
ml-explore	mlx	< 0.29.4

Weakness: CWE-122 · Vector: CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:L/VI:N/VA:L/SC:N/SI:N/SA:N/E:P

References

https://github.com/ml-explore/mlx/security/advisories/GHSA-w6vg-jg77-2qg6 x_refsource_CONFIRM
https://github.com/ml-explore/mlx/pull/1 x_refsource_MISC
https://github.com/ml-explore/mlx/pull/2 x_refsource_MISC