vLLM before 0.29.0 Denial of Service via out-of-range stop_token_ids
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).
CVE-2026-100654
Primary source · Revision 1
Imported from structured advisory data. Source references identify upstream records or fields; verify affected versions and your deployed configuration.
Security context
Source reference: cve:CVE-2026-100654
vllm
Source reference: /containers/cna/affected/0/product
- Affected versions
- Not stated in this record
- Fixed versions
- Not stated in this record
- Prerequisites
- Not stated in this record
Structured affected-version data
Original package identities and version rules from this source.
[
{
"vendor": "vllm-project",
"product": "vllm",
"versions": [
{
"status": "affected",
"version": "0",
"lessThan": "0.29.0",
"versionType": "semver"
},
{
"status": "unaffected",
"version": "0.29.0",
"versionType": "semver"
}
],
"packageURL": "pkg:pypi/vllm",
"defaultStatus": "unaffected"
}
]- Related CVEs
- CVE-2026-100654
Source reference:
/cveMetadata/cveId
Publisher severity
high
Source reference: /containers/cna/metrics
- Event
- #4433 · Revision 1
- Added to Heist Feed
- Publisher date