Vulnerability in Vllm-Project Vllm
CVE-2026-100654
vLLM before 0.29.0 accepts user-controlled stop_token_ids on the OpenAI-compatible POST /v1/completions and POST /v1/chat/completions endpoints but validates only that the values are integers, not that each token id is within the model vocabulary/logits range. When min_tokens > 0, the stop token ids are used as logits indices to suppress stop tokens, so an out-of-range id reaches a CUDA indexing operation (index_put_) and triggers a device-side assertion. An authenticated API user can send a single malformed completion request that returns 500 Internal Server Error and puts EngineCore into a fatal state, causing subsequent requests to fail until the service is restarted (denial of service).
Published .
CVSS v3 metric
CVSS v3 base score 6.5 (Medium). Vector: CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H.
EPSS exploit prediction
No EPSS score from FIRST.org yet. New CVEs usually get one within a few days; no score means no signal yet, not low risk. How EPSS works.
Affected products
Weakness classification (CWE)
References
- github.com/vllm-project/vllm/security/advisories/GHSA-v5gm-qgmv-gc6c (vendor-advisory)
- www.vulncheck.com/advisories/vllm-before-0.29.0-denial-of-service-via-out-of-ra… (third-party-advisory)
Frequently asked questions
- What is CVE-2026-100654?
- CVE-2026-100654 is a medium-severity vulnerability in Vllm-Project Vllm, classified under Improper Validation of Array Index. CVSS score: 6.5/10. Published 2026-09-26.
- How severe is CVE-2026-100654?
- Medium severity. CVSS v3 base score is 6.5 out of 10.