CVE-2025-24357
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.
Does this matter?
High impact if exploited, but EPSS currently rates exploitation as unlikely (0.70%). Schedule it in the normal patch cycle and watch for a rise in EPSS or a public exploit.
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.
- CVSS 3.1
- 8.8 HIGHCVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H
- EPSS
- 0.70% probability · 51th percentile
- CISA KEV
- Not listed
- Weakness
- CWE-502
- Affected
- vllm/vllm
- Source
- security-advisories@github.com
References
- https://github.com/vllm-project/vllm/commit/d3d6bb13fb62da3234addf6574922a4ec0513d04Patch
- https://github.com/vllm-project/vllm/pull/12366Issue Tracking, Patch
- https://github.com/vllm-project/vllm/security/advisories/GHSA-rh4j-5rhw-hr54Vendor Advisory
- https://pytorch.org/docs/stable/generated/torch.load.htmlTechnical Description
Source: NVD record, EPSS from FIRST.org, KEV from CISA. Refreshed daily.