vllm
Affected
- introduced
0.18.0, fixed0.20.0(ecosystem)
Fixed versions
0.20.0
Fix
- Fixed version
0.20.0pip install --upgrade vllm==0.20.0
Advisory
Also known as: CVE-2026-44223, GHSA-83vm-p52w-f9pw
The record has no summary.
vLLM is an inference and serving engine for large language models (LLMs). From to before 0.20.0, the extract_hidden_states speculative decoding proposer in vLLM returns a tensor with an incorrect shape after the first decode step, causing a RuntimeError that crashes the EngineCore process. The crash is triggered when any request in the batch uses sampling penalty parameters (repetition_penalty, frequency_penalty, or presence_penalty). A single request with a penalty parameter (e.g., "repetition_penalty": 1.1) is sufficient to crash the server. This vulnerability is fixed in 0.20.0.
Medium Basis: CVSS v3 score 6.5, vector CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H
0.18.0, fixed 0.20.0 (ecosystem)0.20.0
0.20.0pip install --upgrade vllm==0.20.0Fix commands are the export's own generic commands. Ojuju never runs them. Check them against your setup first.
Check your own computer: guard packages check
Source: PyPA Advisory Database · CC BY 4.0