Vulnerability · Published Jan 27, 2025 · Assigned by GitHub_M

CVE-2025-24357vLLM allows a malicious model RCE by torch.load in hf_model_weights_iterator

High 7.5

In plain English

A high-severity unsafe deserialization in vllm from vllm-project. There are no confirmed reports of exploitation so far. The chance it is exploited in the next 30 days is estimated at 0.9% (higher than 60% of all known vulnerabilities). Fixed versions are available - upgrade the affected package.

Severity (CVSS 3.1)7.5High

Exploit odds (30 days)0.9%top 40.4% of all CVEs

Exploited in the wildNo reportsCISA SSVC: none

Attack toolingNoneno public tooling known

Official description

vLLM is a library for LLM inference and serving. vllm/model_executor/weight_utils.py implements hf_model_weights_iterator to load the model checkpoint, which is downloaded from huggingface. It uses the torch.load function and the weights_only parameter defaults to False. When torch.load loads malicious pickle data, it will execute arbitrary code during unpickling. This vulnerability is fixed in v0.7.0.

Affected packages & fixed versions

Ecosystem	Package	Fixed in	Advisory
PyPI	vllm	0.7.0	GHSA-rh4j-5rhw-hr54
PyPI	vllm	0.7.0	PYSEC-2025-58

Affected products (as reported)

Vendor	Product	Versions
vllm-project	vllm	< 0.7.0

Weakness: CWE-502 · Vector: CVSS:3.1/AV:N/AC:H/PR:N/UI:R/S:U/C:H/I:H/A:H

References

https://github.com/vllm-project/vllm/security/advisories/GHSA-rh4j-5rhw-hr54 x_refsource_CONFIRM
https://github.com/vllm-project/vllm/pull/12366 x_refsource_MISC
https://github.com/vllm-project/vllm/commit/d3d6bb13fb62da3234addf6574922a4ec0513d04 x_refsource_MISC
https://pytorch.org/docs/stable/generated/torch.load.html x_refsource_MISC