UpHeron
All vulnerabilities

Servers and applications

CVE-2026-116181

vLLM 0.20.1 before 0.29.0 DoS and GPU Out-of-Bounds Write via stop_token_ids

Severity
High 7.1
Published
25 Oct 2026
Last changed
29 Oct 2026
Added to UpHeron
3 Nov 2026
Fixed in
0.29.0
Sources
NVD, CVE Program

Description

vLLM from 0.20.1 before 0.29.0 contains an improper input validation vulnerability that allows remote attackers to crash the engine by supplying out-of-vocabulary stop_token_ids. Attackers can send OpenAI chat, completion, generate, or batch requests with invalid IDs to terminate the engine or write outside the V2 runner's logits buffer.

Affected software

SoftwareAffected versions
vLLMServers and applications · cna:vllm-project|vllm
  • From 0.20.1 before 0.29.0
vLLMServers and applications · cpe:a:vllm:vllm
  • From 0.20.1 before 0.29.0

References

  • NVD https://nvd.nist.gov/vuln/detail/CVE-2026-116181
  • CVE.org https://www.cve.org/CVERecord?id=CVE-2026-116181