| Component | vLLM |
| Severity | low (CVSS 3.7) |
| CVSS vector | CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:L/A:N |
| Published (NVD) | 2026-10-21 |
| Affected versions | < 0.31.0 from the CVE Program record written by the CNA that assigned the CVE |
| Fixed in | 0.31.0 |
| Weakness | CWE-440 |
| CISA KEV | not listed |
NVD description
vLLM before 0.31.0 contains a cross-request mask leak in InputBatch.condense() that allows remote attackers to constrain other clients' outputs by leaving stale allowed_token_ids rows. Attackers can submit concurrent constrained generate requests, including via unauthenticated /invocations, so recycled batch slots apply their token allowlist to victims' unrestricted requests.
What to do
- Upgrade vLLM to
0.31.0or later. - Do not expose the component's API or UI to the internet without authentication in front of it.
Sources
- NVD: CVE-2026-114049
- https://github.com/vllm-project/vllm/security/advisories/GHSA-6cxc-2vcg-w5qc
- https://www.vulncheck.com/advisories/vllm-before-0.31.0-stale-allowed-token-ids-mask-via-inputbatch-condense