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.

Classification

CVE ID: CVE-2025-24357

CVSS Base Severity: HIGH

CVSS Base Score: 7.5

Affected Products

Vendor: vllm-project

Product: vllm

Exploit Prediction Scoring System (EPSS)

EPSS Score: 0.05% (probability of being exploited)

EPSS Percentile: 17.96% (scored less or equal to compared to others)

EPSS Date: 2025-02-25 (when was this score calculated)

References

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

Timeline