CVE-2026-100654
Vllm ≤ 0.29.0
Raw vector
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:X/V:X/RE:X/U:XCVSS and EPSS are reproduced from their sources (NVD, FIRST EPSS). EPSS, Forum of Incident Response and Security Teams (FIRST), epss.first.org. Risk Priority is our own derived reading, not an NVD score.
Summary
CVE-2026-100654 is a high-severity Improper Validation of Array Index (CWE-129) vulnerability in Vllm. Its CVSS base score is 7.1 (High).
Operationally, exploitation aligns with the MITRE ATT&CK technique Exploitation for Privilege Escalation (T1068); ranked at the 24th percentile by exploit likelihood (below the median); it is not currently listed in the CISA KEV catalog; a public proof-of-concept is referenced.
This vulnerability is AI-related — categorised as APIs and Models.
OWASP Top 10 for Web (2025)
EU & UK References
- 🇪🇺 ENISA EUVD: EUVD-2026-87748
Vulnerability Data
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…
more
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).
- CWE(s)
AI Security AnalysisAI
- AI Category
- APIs and Models
- Risk Domain
- N/A
- OWASP Top 10 for LLMs 2025
- None mapped
- Classification Reason
- Matched keywords: openai
Related Threats
MITRE ATT&CK Enterprise Techniques
MITRE ATT&CK v20.0
CVEs Like This One
Affected Assets
Mitigating Controls
Mitigating Controls (NIST CSF 2.0)
Our mappings, derived directly from the weakness types (CWEs) cited in the NVD entry via our CWE→CSF cross-walk — links open the control.
Secure SDLC practices directly require input validation and bounds checking that prevent improper array indexing.
Mitigating Controls (ISO/IEC 27001:2022 Annex A)
Our mappings, derived directly from the weakness types (CWEs) cited in the NVD entry via our CWE→ISO cross-walk — links open the control.
Security testing in development can detect out-of-bounds array access but does not prevent the weakness by itself.
Secure coding standards explicitly require bounds checking and validation of array indices derived from untrusted data.