CVE-2026-44223
vLLM: extract_hidden_states speculative decoding crashes server on any request with penalty parameters
Description
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.
How to fix CVE-2026-44223
To remediate CVE-2026-44223, upgrade the affected package to a fixed version below.
- —upgrade to 0.20.0 or later
- —upgrade to 0.20.0 or later
Is CVE-2026-44223 being exploited?
Low — EPSS is 0.0%, meaning exploitation activity has not been observed at scale.
Affected packages (2)
- >= 0.18.0, < 0.20.0
- >= 0.18.0, < 0.20.0
CVSS scores
| Source | Version | Severity | Vector |
|---|---|---|---|
| osv | CVSS 3.1 | MEDIUM6.5 | CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H |