CVE-2026-93841 (GCVE-0-2026-93841)
Vulnerability from cvelistv5 – Published: 2026-09-18 19:06 – Updated: 2026-09-18 19:06
VLAI
EPSS
VEX
Title
vLLM through 0.29.0 Adjacent-Request Sampler State Corruption via Unvalidated Prompt Token IDs
Summary
vLLM through 0.29.0 contains a memory corruption vulnerability in the Triton _bincount_kernel where prompt token IDs index the penalty prompt-presence bitset without bounds checking against vocabulary size. Attackers can submit multimodal audio requests with tokens equal to vocabulary size, causing out-of-bounds writes that corrupt concurrent requests' sampler state and alter repetition penalty behavior.
Severity
CWE
- CWE-129 - Improper Validation of Array Index
Assigner
References
5 references
| URL | Tags |
|---|---|
| https://github.com/vllm-project/vllm/pull/49081 | patchissue-tracking |
| https://github.com/vllm-project/vllm | product |
| https://github.com/vllm-project/vllm/blob/v0.29.0… | technical-description |
| https://github.com/vllm-project/vllm/blob/v0.29.0… | technical-description |
| https://www.vulncheck.com/advisories/vllm-through… | third-party-advisory |
Impacted products
1 product
| Vendor | Product | Version | CPE status | |
|---|---|---|---|---|
| vllm-project | vllm |
Affected:
0 , ≤ 0.29.0
(semver)
cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:* |
Date Public
2026-07-19 00:00
{
"containers": {
"cna": {
"affected": [
{
"defaultStatus": "unaffected",
"packageURL": "pkg:pypi/vllm",
"product": "vllm",
"vendor": "vllm-project",
"versions": [
{
"lessThanOrEqual": "0.29.0",
"status": "affected",
"version": "0",
"versionType": "semver"
}
]
}
],
"cpeApplicability": [
{
"nodes": [
{
"cpeMatch": [
{
"criteria": "cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:*",
"versionEndIncluding": "0.29.0",
"vulnerable": true
}
],
"negate": false,
"operator": "OR"
}
]
}
],
"credits": [
{
"lang": "en",
"type": "reporter",
"value": "Mingkai Yu"
},
{
"lang": "en",
"type": "reporter",
"value": "Jiapeng Li"
},
{
"lang": "en",
"type": "reporter",
"value": "Jiajia Liu"
}
],
"datePublic": "2026-07-19T00:00:00.000Z",
"descriptions": [
{
"lang": "en",
"value": "vLLM through 0.29.0 contains a memory corruption vulnerability in the Triton _bincount_kernel where prompt token IDs index the penalty prompt-presence bitset without bounds checking against vocabulary size. Attackers can submit multimodal audio requests with tokens equal to vocabulary size, causing out-of-bounds writes that corrupt concurrent requests\u0027 sampler state and alter repetition penalty behavior."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "LOW",
"attackRequirements": "PRESENT",
"attackVector": "NETWORK",
"baseScore": 6.3,
"baseSeverity": "MEDIUM",
"privilegesRequired": "NONE",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "NONE",
"vectorString": "CVSS:4.0/AV:N/AC:L/AT:P/PR:N/UI:N/VC:N/VI:L/VA:N/SC:N/SI:N/SA:N",
"version": "4.0",
"vulnAvailabilityImpact": "NONE",
"vulnConfidentialityImpact": "NONE",
"vulnIntegrityImpact": "LOW"
},
"format": "CVSS"
},
{
"cvssV3_1": {
"attackComplexity": "HIGH",
"attackVector": "NETWORK",
"availabilityImpact": "NONE",
"baseScore": 3.7,
"baseSeverity": "LOW",
"confidentialityImpact": "NONE",
"integrityImpact": "LOW",
"privilegesRequired": "NONE",
"scope": "UNCHANGED",
"userInteraction": "NONE",
"vectorString": "CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:L/A:N",
"version": "3.1"
},
"format": "CVSS"
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-129",
"description": "Improper Validation of Array Index",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-09-18T19:06:06.636Z",
"orgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"shortName": "VulnCheck"
},
"references": [
{
"tags": [
"patch",
"issue-tracking"
],
"url": "https://github.com/vllm-project/vllm/pull/49081"
},
{
"tags": [
"product"
],
"url": "https://github.com/vllm-project/vllm"
},
{
"tags": [
"technical-description"
],
"url": "https://github.com/vllm-project/vllm/blob/v0.29.0/vllm/v1/worker/gpu/sample/penalties.py#L241-L255"
},
{
"tags": [
"technical-description"
],
"url": "https://github.com/vllm-project/vllm/blob/v0.29.0/vllm/v1/worker/gpu/sample/penalties.py#L33-L37"
},
{
"name": "VulnCheck Advisory: vLLM through 0.29.0 Adjacent-Request Sampler State Corruption via Unvalidated Prompt Token IDs",
"tags": [
"third-party-advisory"
],
"url": "https://www.vulncheck.com/advisories/vllm-through-0.29.0-adjacent-request-sampler-state-corruption-via-unvalidated-prompt-token-ids"
}
],
"title": "vLLM through 0.29.0 Adjacent-Request Sampler State Corruption via Unvalidated Prompt Token IDs",
"x_generator": {
"engine": "vulncheck-endgame"
}
}
},
"cveMetadata": {
"assignerOrgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"assignerShortName": "VulnCheck",
"cveId": "CVE-2026-93841",
"datePublished": "2026-09-18T19:06:06.636Z",
"dateReserved": "2026-09-18T18:16:52.304Z",
"dateUpdated": "2026-09-18T19:06:06.636Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2",
"vulnerability-lookup:meta": {
"nvd": {
"cve": {
"affected": [
{
"affectedData": [
{
"defaultStatus": "unaffected",
"packageURL": "pkg:pypi/vllm",
"product": "vllm",
"vendor": "vllm-project",
"versions": [
{
"lessThanOrEqual": "0.29.0",
"status": "affected",
"version": "0",
"versionType": "semver"
}
]
}
],
"source": "disclosure@vulncheck.com"
}
],
"cveTags": [],
"descriptions": [
{
"lang": "en",
"value": "vLLM through 0.29.0 contains a memory corruption vulnerability in the Triton _bincount_kernel where prompt token IDs index the penalty prompt-presence bitset without bounds checking against vocabulary size. Attackers can submit multimodal audio requests with tokens equal to vocabulary size, causing out-of-bounds writes that corrupt concurrent requests\u0027 sampler state and alter repetition penalty behavior."
}
],
"id": "CVE-2026-93841",
"lastModified": "2026-09-18T20:17:34.357",
"metrics": {
"cvssMetricV31": [
{
"cvssData": {
"attackComplexity": "HIGH",
"attackVector": "NETWORK",
"availabilityImpact": "NONE",
"baseScore": 3.7,
"baseSeverity": "LOW",
"confidentialityImpact": "NONE",
"integrityImpact": "LOW",
"privilegesRequired": "NONE",
"scope": "UNCHANGED",
"userInteraction": "NONE",
"vectorString": "CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:L/A:N",
"version": "3.1"
},
"exploitabilityScore": 2.2,
"impactScore": 1.4,
"source": "disclosure@vulncheck.com",
"type": "Primary"
}
],
"cvssMetricV40": [
{
"cvssData": {
"Automatable": "NOT_DEFINED",
"Recovery": "NOT_DEFINED",
"Safety": "NOT_DEFINED",
"attackComplexity": "LOW",
"attackRequirements": "PRESENT",
"attackVector": "NETWORK",
"availabilityRequirement": "NOT_DEFINED",
"baseScore": 6.3,
"baseSeverity": "MEDIUM",
"confidentialityRequirement": "NOT_DEFINED",
"exploitMaturity": "NOT_DEFINED",
"integrityRequirement": "NOT_DEFINED",
"modifiedAttackComplexity": "NOT_DEFINED",
"modifiedAttackRequirements": "NOT_DEFINED",
"modifiedAttackVector": "NOT_DEFINED",
"modifiedPrivilegesRequired": "NOT_DEFINED",
"modifiedSubAvailabilityImpact": "NOT_DEFINED",
"modifiedSubConfidentialityImpact": "NOT_DEFINED",
"modifiedSubIntegrityImpact": "NOT_DEFINED",
"modifiedUserInteraction": "NOT_DEFINED",
"modifiedVulnAvailabilityImpact": "NOT_DEFINED",
"modifiedVulnConfidentialityImpact": "NOT_DEFINED",
"modifiedVulnIntegrityImpact": "NOT_DEFINED",
"privilegesRequired": "NONE",
"providerUrgency": "NOT_DEFINED",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "NONE",
"valueDensity": "NOT_DEFINED",
"vectorString": "CVSS:4.0/AV:N/AC:L/AT:P/PR:N/UI:N/VC:N/VI:L/VA:N/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:X",
"version": "4.0",
"vulnAvailabilityImpact": "NONE",
"vulnConfidentialityImpact": "NONE",
"vulnIntegrityImpact": "LOW",
"vulnerabilityResponseEffort": "NOT_DEFINED"
},
"source": "disclosure@vulncheck.com",
"type": "Secondary"
}
]
},
"published": "2026-09-18T20:17:34.357",
"references": [
{
"source": "disclosure@vulncheck.com",
"url": "https://github.com/vllm-project/vllm"
},
{
"source": "disclosure@vulncheck.com",
"url": "https://github.com/vllm-project/vllm/blob/v0.29.0/vllm/v1/worker/gpu/sample/penalties.py#L241-L255"
},
{
"source": "disclosure@vulncheck.com",
"url": "https://github.com/vllm-project/vllm/blob/v0.29.0/vllm/v1/worker/gpu/sample/penalties.py#L33-L37"
},
{
"source": "disclosure@vulncheck.com",
"url": "https://github.com/vllm-project/vllm/pull/49081"
},
{
"source": "disclosure@vulncheck.com",
"url": "https://www.vulncheck.com/advisories/vllm-through-0.29.0-adjacent-request-sampler-state-corruption-via-unvalidated-prompt-token-ids"
}
],
"sourceIdentifier": "disclosure@vulncheck.com",
"vulnStatus": "Received",
"weaknesses": [
{
"description": [
{
"lang": "en",
"value": "CWE-129"
}
],
"source": "disclosure@vulncheck.com",
"type": "Primary"
}
]
}
}
}
}
Loading…
Loading…
Experimental. This forecast is provided for visualization only and may change without notice. Do not use it for operational decisions.
Forecast uses a logistic model when the trend is rising, or an exponential decay model when the trend is falling. Fitted via linearized least squares.
Sightings
| Author | Source | Type | Date | Other |
|---|
Nomenclature
- Seen: The vulnerability was mentioned, discussed, or observed by the user.
- Confirmed: The vulnerability has been validated from an analyst's perspective.
- Published Proof of Concept: A public proof of concept is available for this vulnerability.
- Exploited: The vulnerability was observed as exploited by the user who reported the sighting.
- Patched: The vulnerability was observed as successfully patched by the user who reported the sighting.
- Not exploited: The vulnerability was not observed as exploited by the user who reported the sighting.
- Not confirmed: The user expressed doubt about the validity of the vulnerability.
- Not patched: The vulnerability was not observed as successfully patched by the user who reported the sighting.
Loading…
The MITRE ATT&CK techniques below are AI-generated suggestions, inferred from the description of the
vulnerability by the CIRCL/vulnerability-attack-technique-classification-roberta-base
model, served locally by ML-Gateway.
They have not been verified by an analyst and are provided for guidance only.
The approach is described in our paper Mapping CVEs to MITRE ATT&CK Techniques: A Curated Gold-Set Classifier and the Limits of LLM-Assisted Label Expansion.
Browse all ATT&CK techniques and the vulnerabilities related to each.
The approach is described in our paper Mapping CVEs to MITRE ATT&CK Techniques: A Curated Gold-Set Classifier and the Limits of LLM-Assisted Label Expansion.
Browse all ATT&CK techniques and the vulnerabilities related to each.
Loading…
Related by attack behaviour
Vulnerabilities whose description is nearest to this one in the vector space of the CIRCL/vulnerability-attack-technique-biencoder model. This is a similarity search over the bi-encoder space (plain cosine), not a classification, and it has no measured accuracy.
Loading…