← Vulnerability feed

Vulnerability record · CVE-2026-100654 · published 26 September 2026

CVE-2026-100654: vLLM before 0.29.0 accepts user-controlled stop_token_ids on the OpenAI-compatible POST /v1/completions and PO…

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).

7.1 CVSS 4.0 High EPSS 0.31% · top 79.0% CWE-129 · CWE-129 Received
7.1CVSS 4.0 base score
0.31%EPSS exploitation probability, 30 days
NoNot in CISA KEV
0Affected product versions listed by NVD
2References
26 Sep 2026Last modified by NVD

Description

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).

CVSS:4.0/AV:N/AC:L/AT:N/PR:L/UI:N/VC:N/VI:N/VA:H/SC:N/SI:N/SA:N/E:X/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/MVI:X/MVA:X/MSC:X/MSI:X/MSA:X/S:X/AU:X/R:

References

LinkTags
https://github.com/vllm-project/vllm/security/advisories/GHSA-v5gm-qgmv-gc6c
https://www.vulncheck.com/advisories/vllm-before-0.29.0-denial-of-service-via-out-of-range-stop-token-ids

Track CVE-2026-100654 inside VULONE

Watch it alongside the ransomware crews, C2 infrastructure and forum chatter that reference it, query it through the API and pull it into your SIEM over TAXII.

Start free Open in platform

Related vulnerabilities

Same products first, then exploited flaws of the same weakness class.

8.8CVE-2022-48503Apple WebKit bounds check flaw allows code execution via web contentApple fixed an insufficient bounds-checking issue in WebKit across Safari, iOS, iPadOS, macOS, tvOS and watchOS. Processing malicious web content can…KEVEPSS 3.2%analysed

Source: NIST National Vulnerability Database (record CVE-2026-100654), CISA KEV, FIRST EPSS (scores of 2026-09-27). This page is refreshed as NVD updates the record.