Search
Find a vulnerability
Search criteria
176 vulnerabilities by vllm-project
CVE-2026-103241 (GCVE-0-2026-103241)
Vulnerability from nvd – Published: 2026-09-30 16:45 – Updated: 2026-09-30 16:45 X_Open Source
VLAI
EPSS
VEX
Title
vllm-project vLLM Gemma4UnifiedParser gemma4.rs denial of service
Summary
A flaw has been found in vllm-project vLLM up to 0.26.0. This vulnerability affects unknown code of the file rust/src/parser/src/unified/gemma4.rs of the component Gemma4UnifiedParser. Executing a manipulation can lead to denial of service. The attack may be launched remotely. The exploit has been published and may be used. Upgrading to version 0.29.1rc0 is able to resolve this issue. This patch is called 3439bad37e68ba9755a46f4f6b44a4aeaf1f60a9. Upgrading the affected component is advised.
Severity
CWE
- CWE-404 - Denial of Service
Assigner
References
10 references
| URL | Tags |
|---|---|
| https://vuldb.com/vuln/411965 | vdb-entry |
| https://vuldb.com/vuln/411965/cti | signaturepermissions-required |
| https://vuldb.com/cve/CVE-2026-103241 | third-party-advisory |
| https://vuldb.com/submit/956250 | third-party-advisory |
| https://github.com/vllm-project/vllm/issues/50927 | issue-tracking |
| https://github.com/vllm-project/vllm/pull/54303 | issue-trackingpatch |
| https://gist.github.com/Yunzez/8e98d656aa667095b5… | exploit |
| https://github.com/vllm-project/vllm/commit/3439b… | patch |
| https://github.com/vllm-project/vllm/releases/tag… | patch |
| https://github.com/vllm-project/vllm/ | product |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| vllm-project | vLLM |
Affected:
0.1
Affected: 0.2 Affected: 0.3 Affected: 0.4 Affected: 0.5 Affected: 0.6 Affected: 0.7 Affected: 0.8 Affected: 0.9 Affected: 0.10 Affected: 0.11 Affected: 0.12 Affected: 0.13 Affected: 0.14 Affected: 0.15 Affected: 0.16 Affected: 0.17 Affected: 0.18 Affected: 0.19 Affected: 0.20 Affected: 0.21 Affected: 0.22 Affected: 0.23 Affected: 0.24 Affected: 0.25 Affected: 0.26.0 Unaffected: 0.29.1rc0 cpe:2.3:a:vllm-project:vllm:*:*:*:*:*:*:*:* |
{
"containers": {
"cna": {
"affected": [
{
"cpes": [
"cpe:2.3:a:vllm-project:vllm:*:*:*:*:*:*:*:*"
],
"modules": [
"Gemma4UnifiedParser"
],
"product": "vLLM",
"vendor": "vllm-project",
"versions": [
{
"status": "affected",
"version": "0.1"
},
{
"status": "affected",
"version": "0.2"
},
{
"status": "affected",
"version": "0.3"
},
{
"status": "affected",
"version": "0.4"
},
{
"status": "affected",
"version": "0.5"
},
{
"status": "affected",
"version": "0.6"
},
{
"status": "affected",
"version": "0.7"
},
{
"status": "affected",
"version": "0.8"
},
{
"status": "affected",
"version": "0.9"
},
{
"status": "affected",
"version": "0.10"
},
{
"status": "affected",
"version": "0.11"
},
{
"status": "affected",
"version": "0.12"
},
{
"status": "affected",
"version": "0.13"
},
{
"status": "affected",
"version": "0.14"
},
{
"status": "affected",
"version": "0.15"
},
{
"status": "affected",
"version": "0.16"
},
{
"status": "affected",
"version": "0.17"
},
{
"status": "affected",
"version": "0.18"
},
{
"status": "affected",
"version": "0.19"
},
{
"status": "affected",
"version": "0.20"
},
{
"status": "affected",
"version": "0.21"
},
{
"status": "affected",
"version": "0.22"
},
{
"status": "affected",
"version": "0.23"
},
{
"status": "affected",
"version": "0.24"
},
{
"status": "affected",
"version": "0.25"
},
{
"status": "affected",
"version": "0.26.0"
},
{
"status": "unaffected",
"version": "0.29.1rc0"
}
]
}
],
"credits": [
{
"lang": "en",
"type": "reporter",
"value": "Zyz3366 (VulDB User)"
}
],
"descriptions": [
{
"lang": "en",
"value": "A flaw has been found in vllm-project vLLM up to 0.26.0. This vulnerability affects unknown code of the file rust/src/parser/src/unified/gemma4.rs of the component Gemma4UnifiedParser. Executing a manipulation can lead to denial of service. The attack may be launched remotely. The exploit has been published and may be used. Upgrading to version 0.29.1rc0 is able to resolve this issue. This patch is called 3439bad37e68ba9755a46f4f6b44a4aeaf1f60a9. Upgrading the affected component is advised."
}
],
"metrics": [
{
"cvssV4_0": {
"baseScore": 6.9,
"baseSeverity": "MEDIUM",
"vectorString": "CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:N/VI:N/VA:L/SC:N/SI:N/SA:N/E:P",
"version": "4.0"
}
},
{
"cvssV3_1": {
"baseScore": 5.3,
"baseSeverity": "MEDIUM",
"vectorString": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:L/E:P/RL:O/RC:C",
"version": "3.1"
}
},
{
"cvssV3_0": {
"baseScore": 5.3,
"baseSeverity": "MEDIUM",
"vectorString": "CVSS:3.0/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:L/E:P/RL:O/RC:C",
"version": "3.0"
}
},
{
"cvssV2_0": {
"baseScore": 5,
"vectorString": "AV:N/AC:L/Au:N/C:N/I:N/A:P/E:POC/RL:OF/RC:C",
"version": "2.0"
}
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-404",
"description": "Denial of Service",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-09-30T16:45:13.581Z",
"orgId": "1af790b2-7ee1-4545-860a-a788eba489b5",
"shortName": "VulDB"
},
"references": [
{
"name": "VDB-411965 | vllm-project vLLM Gemma4UnifiedParser gemma4.rs denial of service",
"tags": [
"vdb-entry"
],
"url": "https://vuldb.com/vuln/411965"
},
{
"name": "VDB-411965 | CTI Indicators (IOB, IOC, TTP, IOA)",
"tags": [
"signature",
"permissions-required"
],
"url": "https://vuldb.com/vuln/411965/cti"
},
{
"name": "CVE-2026-103241 | CVE Analysis and Report",
"tags": [
"third-party-advisory"
],
"url": "https://vuldb.com/cve/CVE-2026-103241"
},
{
"name": "Submit #956250 | vLLM Project vLLM v0.26.0 Denial of Service",
"tags": [
"third-party-advisory"
],
"url": "https://vuldb.com/submit/956250"
},
{
"tags": [
"issue-tracking"
],
"url": "https://github.com/vllm-project/vllm/issues/50927"
},
{
"tags": [
"issue-tracking",
"patch"
],
"url": "https://github.com/vllm-project/vllm/pull/54303"
},
{
"tags": [
"exploit"
],
"url": "https://gist.github.com/Yunzez/8e98d656aa667095b513161eb056d28e"
},
{
"tags": [
"patch"
],
"url": "https://github.com/vllm-project/vllm/commit/3439bad37e68ba9755a46f4f6b44a4aeaf1f60a9"
},
{
"tags": [
"patch"
],
"url": "https://github.com/vllm-project/vllm/releases/tag/v0.29.1rc0"
},
{
"tags": [
"product"
],
"url": "https://github.com/vllm-project/vllm/"
}
],
"tags": [
"x_open-source"
],
"timeline": [
{
"lang": "en",
"time": "2026-09-30T00:00:00.000Z",
"value": "Advisory disclosed"
},
{
"lang": "en",
"time": "2026-09-30T02:00:00.000Z",
"value": "VulDB entry created"
},
{
"lang": "en",
"time": "2026-09-30T12:42:01.000Z",
"value": "VulDB entry last update"
}
],
"title": "vllm-project vLLM Gemma4UnifiedParser gemma4.rs denial of service",
"x_generator": [
"VulDB PVTS v202609"
]
}
},
"cveMetadata": {
"assignerOrgId": "1af790b2-7ee1-4545-860a-a788eba489b5",
"assignerShortName": "VulDB",
"cveId": "CVE-2026-103241",
"datePublished": "2026-09-30T16:45:13.581Z",
"dateReserved": "2026-09-30T10:36:13.164Z",
"dateUpdated": "2026-09-30T16:45:13.581Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-100654 (GCVE-0-2026-100654)
Vulnerability from nvd – Published: 2026-09-26 13:23 – Updated: 2026-09-30 14:51
VLAI
EPSS
VEX
Title
vLLM before 0.29.0 Denial of Service via out-of-range stop_token_ids
Summary
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).
Severity
SSVC
Exploitation: none
Automatable: no
Technical Impact: partial
CISA Coordinator · CISA-ADP (v2.0.3)
Decision recorded 2026-09-30 14:51 UTC
CWE
- CWE-129 - Improper Validation of Array Index
Assigner
References
2 references
| URL | Tags |
|---|---|
| https://github.com/vllm-project/vllm/security/adv… | vendor-advisory |
| https://www.vulncheck.com/advisories/vllm-before-… | third-party-advisory |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| vllm-project | vllm |
Affected:
0 , < 0.29.0
(semver)
Unaffected: 0.29.0 (semver) cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:* |
Date Public
2026-09-12 00:00
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-100654",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "no"
},
{
"Technical Impact": "partial"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-09-30T14:51:29.949766Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-09-30T14:51:39.132Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"cna": {
"affected": [
{
"defaultStatus": "unaffected",
"packageURL": "pkg:pypi/vllm",
"product": "vllm",
"vendor": "vllm-project",
"versions": [
{
"lessThan": "0.29.0",
"status": "affected",
"version": "0",
"versionType": "semver"
},
{
"status": "unaffected",
"version": "0.29.0",
"versionType": "semver"
}
]
}
],
"cpeApplicability": [
{
"nodes": [
{
"cpeMatch": [
{
"criteria": "cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:*",
"versionEndExcluding": "0.29.0",
"vulnerable": true
}
],
"negate": false,
"operator": "OR"
}
]
}
],
"credits": [
{
"lang": "en",
"type": "reporter",
"value": "passer12"
},
{
"lang": "en",
"type": "coordinator",
"value": "jperezdealgaba"
},
{
"lang": "en",
"type": "finder",
"value": "QwertyJack"
}
],
"datePublic": "2026-09-12T00:00:00.000Z",
"descriptions": [
{
"lang": "en",
"value": "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 \u003e 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)."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "LOW",
"attackRequirements": "NONE",
"attackVector": "NETWORK",
"baseScore": 7.1,
"baseSeverity": "HIGH",
"privilegesRequired": "LOW",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "NONE",
"vectorString": "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",
"version": "4.0",
"vulnAvailabilityImpact": "HIGH",
"vulnConfidentialityImpact": "NONE",
"vulnIntegrityImpact": "NONE"
},
"format": "CVSS"
},
{
"cvssV3_1": {
"attackComplexity": "LOW",
"attackVector": "NETWORK",
"availabilityImpact": "HIGH",
"baseScore": 6.5,
"baseSeverity": "MEDIUM",
"confidentialityImpact": "NONE",
"integrityImpact": "NONE",
"privilegesRequired": "LOW",
"scope": "UNCHANGED",
"userInteraction": "NONE",
"vectorString": "CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H",
"version": "3.1"
},
"format": "CVSS"
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-129",
"description": "Improper Validation of Array Index",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-09-26T13:23:23.312Z",
"orgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"shortName": "VulnCheck"
},
"references": [
{
"name": "GitHub Security Advisory (GHSA-v5gm-qgmv-gc6c)",
"tags": [
"vendor-advisory"
],
"url": "https://github.com/vllm-project/vllm/security/advisories/GHSA-v5gm-qgmv-gc6c"
},
{
"name": "VulnCheck Advisory: vLLM before 0.29.0 Denial of Service via out-of-range stop_token_ids",
"tags": [
"third-party-advisory"
],
"url": "https://www.vulncheck.com/advisories/vllm-before-0.29.0-denial-of-service-via-out-of-range-stop-token-ids"
}
],
"title": "vLLM before 0.29.0 Denial of Service via out-of-range stop_token_ids",
"x_generator": {
"engine": "vulncheck-endgame"
}
}
},
"cveMetadata": {
"assignerOrgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"assignerShortName": "VulnCheck",
"cveId": "CVE-2026-100654",
"datePublished": "2026-09-26T13:23:23.312Z",
"dateReserved": "2026-09-26T02:33:07.899Z",
"dateUpdated": "2026-09-30T14:51:39.132Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-100653 (GCVE-0-2026-100653)
Vulnerability from nvd – Published: 2026-09-26 13:23 – Updated: 2026-10-01 11:06
VLAI
EPSS
VEX
Title
vLLM 0.22.1 before 0.28.0 Incomplete Artifact Pin Propagation
Summary
vLLM is an inference and serving engine for large language models. In versions from 0.22.1 through 0.28.0, the operator-supplied model revision pin (--revision / --code-revision) is not propagated to several Hugging Face artifact loads for the FunAudioChat and Tarsier2 architectures: the WhisperFeatureExtractor and speech_tokenizer PreTrainedTokenizerFast loads in vllm/model_executor/models/funaudiochat.py and the Qwen2VLConfig.from_pretrained call used by Tarsier2ProcessingInfo in vllm/model_executor/models/qwen2_vl.py. As a result, deployments pinned to a reviewed revision still resolve these behavior-affecting processor, tokenizer, and config artifacts from the repository's default revision, so a later change to the upstream default branch can alter audio preprocessing, speech tokenizer behavior, or Tarsier2 configuration without any change to the operator's configured pin. This is a supply-chain integrity and reproducibility failure for pinned deployments; it is residual to the earlier fix tracked as GHSA-3ww4-5jv9-j5gm / CVE-2026-47155 and does not constitute remote code execution or a trust_remote_code=False bypass. The issue is fixed in version 0.28.0.
Severity
SSVC
Exploitation: poc
Automatable: no
Technical Impact: partial
CISA Coordinator · CISA-ADP (v2.0.3)
Decision recorded 2026-09-30 17:46 UTC
CWE
- CWE-348 - Use of Less Trusted Source
Assigner
References
3 references
| URL | Tags |
|---|---|
| https://github.com/vllm-project/vllm/security/adv… | vendor-advisory |
| https://github.com/vllm-project/vllm/commit/d26a2… | patch |
| https://www.vulncheck.com/advisories/vllm-0.22.1-… | third-party-advisory |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| vllm-project | vllm |
Affected:
0.22.1 , < 0.28.0
(semver)
Unaffected: 0.28.0 (semver) cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:* |
Date Public
2026-09-12 00:00
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-100653",
"options": [
{
"Exploitation": "poc"
},
{
"Automatable": "no"
},
{
"Technical Impact": "partial"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-09-30T17:46:29.258369Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-09-30T17:46:36.993Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"references": [
{
"tags": [
"exploit"
],
"url": "https://github.com/vllm-project/vllm/security/advisories/GHSA-hhv2-872h-628q"
}
],
"title": "CISA ADP Vulnrichment"
}
],
"cna": {
"affected": [
{
"defaultStatus": "unaffected",
"packageURL": "pkg:pypi/vllm",
"product": "vllm",
"vendor": "vllm-project",
"versions": [
{
"lessThan": "0.28.0",
"status": "affected",
"version": "0.22.1",
"versionType": "semver"
},
{
"status": "unaffected",
"version": "0.28.0",
"versionType": "semver"
}
]
}
],
"cpeApplicability": [
{
"nodes": [
{
"cpeMatch": [
{
"criteria": "cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:*",
"versionEndExcluding": "0.28.0",
"versionStartIncluding": "0.22.1",
"vulnerable": true
}
],
"negate": false,
"operator": "OR"
}
]
}
],
"credits": [
{
"lang": "en",
"type": "reporter",
"value": "rexpository"
},
{
"lang": "en",
"type": "coordinator",
"value": "jperezdealgaba"
},
{
"lang": "en",
"type": "remediation developer",
"value": "KernelClint"
}
],
"datePublic": "2026-09-12T00:00:00.000Z",
"descriptions": [
{
"lang": "en",
"value": "vLLM is an inference and serving engine for large language models. In versions from 0.22.1 through 0.28.0, the operator-supplied model revision pin (--revision / --code-revision) is not propagated to several Hugging Face artifact loads for the FunAudioChat and Tarsier2 architectures: the WhisperFeatureExtractor and speech_tokenizer PreTrainedTokenizerFast loads in vllm/model_executor/models/funaudiochat.py and the Qwen2VLConfig.from_pretrained call used by Tarsier2ProcessingInfo in vllm/model_executor/models/qwen2_vl.py. As a result, deployments pinned to a reviewed revision still resolve these behavior-affecting processor, tokenizer, and config artifacts from the repository\u0027s default revision, so a later change to the upstream default branch can alter audio preprocessing, speech tokenizer behavior, or Tarsier2 configuration without any change to the operator\u0027s configured pin. This is a supply-chain integrity and reproducibility failure for pinned deployments; it is residual to the earlier fix tracked as GHSA-3ww4-5jv9-j5gm / CVE-2026-47155 and does not constitute remote code execution or a trust_remote_code=False bypass. The issue is fixed in version 0.28.0."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "HIGH",
"attackRequirements": "PRESENT",
"attackVector": "NETWORK",
"baseScore": 8.3,
"baseSeverity": "HIGH",
"privilegesRequired": "NONE",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "NONE",
"vectorString": "CVSS:4.0/AV:N/AC:H/AT:P/PR:N/UI:N/VC:L/VI:H/VA:N/SC:N/SI:N/SA:N",
"version": "4.0",
"vulnAvailabilityImpact": "NONE",
"vulnConfidentialityImpact": "LOW",
"vulnIntegrityImpact": "HIGH"
},
"format": "CVSS"
},
{
"cvssV3_1": {
"attackComplexity": "HIGH",
"attackVector": "NETWORK",
"availabilityImpact": "NONE",
"baseScore": 6.5,
"baseSeverity": "MEDIUM",
"confidentialityImpact": "LOW",
"integrityImpact": "HIGH",
"privilegesRequired": "NONE",
"scope": "UNCHANGED",
"userInteraction": "NONE",
"vectorString": "CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:L/I:H/A:N",
"version": "3.1"
},
"format": "CVSS"
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-348",
"description": "Use of Less Trusted Source",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-10-01T11:06:45.139Z",
"orgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"shortName": "VulnCheck"
},
"references": [
{
"name": "GitHub Security Advisory (GHSA-hhv2-872h-628q)",
"tags": [
"vendor-advisory"
],
"url": "https://github.com/vllm-project/vllm/security/advisories/GHSA-hhv2-872h-628q"
},
{
"name": "Patch Commit",
"tags": [
"patch"
],
"url": "https://github.com/vllm-project/vllm/commit/d26a28ab033697f55a1414b5b0435de7cd6045b6"
},
{
"name": "VulnCheck Advisory: vLLM 0.22.1 before 0.28.0 Incomplete Artifact Pin Propagation",
"tags": [
"third-party-advisory"
],
"url": "https://www.vulncheck.com/advisories/vllm-0.22.1-before-0.28.0-incomplete-artifact-pin-propagation"
}
],
"title": "vLLM 0.22.1 before 0.28.0 Incomplete Artifact Pin Propagation",
"x_generator": {
"engine": "vulncheck-endgame"
}
}
},
"cveMetadata": {
"assignerOrgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"assignerShortName": "VulnCheck",
"cveId": "CVE-2026-100653",
"datePublished": "2026-09-26T13:23:22.605Z",
"dateReserved": "2026-09-26T02:33:07.899Z",
"dateUpdated": "2026-10-01T11:06:45.139Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-100652 (GCVE-0-2026-100652)
Vulnerability from nvd – Published: 2026-09-26 13:23 – Updated: 2026-09-28 16:42
VLAI
EPSS
VEX
Title
vLLM 0.22.0 through 0.23.0 Denial of Service via stop_token_ids
Summary
vLLM versions 0.22.0 through 0.23.0 fail to validate stop_token_ids against vocabulary bounds in Rust HTTP and gRPC frontends, allowing out-of-vocabulary token IDs to reach MinTokensLogitsProcessor. Attackers can submit requests with min_tokens greater than zero and out-of-vocabulary stop_token_ids to trigger CUDA tensor indexing failures that leave EngineCore in a fatal state requiring service restart.
Severity
SSVC
Exploitation: none
Automatable: no
Technical Impact: partial
CISA Coordinator · CISA-ADP (v2.0.3)
Decision recorded 2026-09-28 16:41 UTC
CWE
- CWE-20 - Improper Input Validation
Assigner
References
2 references
| URL | Tags |
|---|---|
| https://github.com/vllm-project/vllm/security/adv… | vendor-advisory |
| https://www.vulncheck.com/advisories/vllm-0.22.0-… | third-party-advisory |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| vllm-project | vllm |
Affected:
0.22.0 , < 0.24.0
(semver)
Unaffected: 0.24.0 (semver) cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:* |
Date Public
2026-09-12 00:00
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-100652",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "no"
},
{
"Technical Impact": "partial"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-09-28T16:41:47.920735Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-09-28T16:42:13.668Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"cna": {
"affected": [
{
"defaultStatus": "unaffected",
"packageURL": "pkg:pypi/vllm",
"product": "vllm",
"vendor": "vllm-project",
"versions": [
{
"lessThan": "0.24.0",
"status": "affected",
"version": "0.22.0",
"versionType": "semver"
},
{
"status": "unaffected",
"version": "0.24.0",
"versionType": "semver"
}
]
}
],
"cpeApplicability": [
{
"nodes": [
{
"cpeMatch": [
{
"criteria": "cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:*",
"versionEndExcluding": "0.24.0",
"versionStartIncluding": "0.22.0",
"vulnerable": true
}
],
"negate": false,
"operator": "OR"
}
]
}
],
"credits": [
{
"lang": "en",
"type": "reporter",
"value": "rexpository"
},
{
"lang": "en",
"type": "coordinator",
"value": "jperezdealgaba"
}
],
"datePublic": "2026-09-12T00:00:00.000Z",
"descriptions": [
{
"lang": "en",
"value": "vLLM versions 0.22.0 through 0.23.0 fail to validate stop_token_ids against vocabulary bounds in Rust HTTP and gRPC frontends, allowing out-of-vocabulary token IDs to reach MinTokensLogitsProcessor. Attackers can submit requests with min_tokens greater than zero and out-of-vocabulary stop_token_ids to trigger CUDA tensor indexing failures that leave EngineCore in a fatal state requiring service restart."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "HIGH",
"attackRequirements": "PRESENT",
"attackVector": "NETWORK",
"baseScore": 8.2,
"baseSeverity": "HIGH",
"privilegesRequired": "NONE",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "NONE",
"vectorString": "CVSS:4.0/AV:N/AC:H/AT:P/PR:N/UI:N/VC:N/VI:N/VA:H/SC:N/SI:N/SA:N",
"version": "4.0",
"vulnAvailabilityImpact": "HIGH",
"vulnConfidentialityImpact": "NONE",
"vulnIntegrityImpact": "NONE"
},
"format": "CVSS"
},
{
"cvssV3_1": {
"attackComplexity": "HIGH",
"attackVector": "NETWORK",
"availabilityImpact": "HIGH",
"baseScore": 5.9,
"baseSeverity": "MEDIUM",
"confidentialityImpact": "NONE",
"integrityImpact": "NONE",
"privilegesRequired": "NONE",
"scope": "UNCHANGED",
"userInteraction": "NONE",
"vectorString": "CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:N/A:H",
"version": "3.1"
},
"format": "CVSS"
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-20",
"description": "Improper Input Validation",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-09-26T13:23:21.865Z",
"orgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"shortName": "VulnCheck"
},
"references": [
{
"name": "GitHub Security Advisory (GHSA-qff2-492f-9fm4)",
"tags": [
"vendor-advisory"
],
"url": "https://github.com/vllm-project/vllm/security/advisories/GHSA-qff2-492f-9fm4"
},
{
"name": "VulnCheck Advisory: vLLM 0.22.0 through 0.23.0 Denial of Service via stop_token_ids",
"tags": [
"third-party-advisory"
],
"url": "https://www.vulncheck.com/advisories/vllm-0.22.0-through-0.23.0-denial-of-service-via-stop-token-ids"
}
],
"title": "vLLM 0.22.0 through 0.23.0 Denial of Service via stop_token_ids",
"x_generator": {
"engine": "vulncheck-endgame"
}
}
},
"cveMetadata": {
"assignerOrgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"assignerShortName": "VulnCheck",
"cveId": "CVE-2026-100652",
"datePublished": "2026-09-26T13:23:21.865Z",
"dateReserved": "2026-09-26T02:33:07.899Z",
"dateUpdated": "2026-09-28T16:42:13.668Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-100651 (GCVE-0-2026-100651)
Vulnerability from nvd – Published: 2026-09-26 13:23 – Updated: 2026-09-28 18:25
VLAI
EPSS
VEX
Title
vllm before 0.29.0 Denial of Service via Decoder Prompt Length Bypass
Summary
vLLM before 0.29.0 fails to enforce decoder prompt-length validation on the disaggregated serving endpoint /inference/v1/generate. When the request contains a 'features' (multimodal) payload, vllm/entrypoints/serve/disagg/serving.py builds a multimodal EngineInput directly from the caller-supplied token_ids, and GenerateRequest.token_ids (vllm/entrypoints/serve/disagg/protocol.py) is not checked against model_config.max_model_len. For multimodal processors that report skip_prompt_length_check=True (for example Nemotron Parse, Whisper, and FireRedLID), InputProcessor._validate_prompt_len() returns immediately for both encoder and decoder prompts, so an overlong prompt becomes an EngineCoreRequest and reaches the worker input-batch copy into a fixed max_model_len-wide NumPy row. A client able to reach the endpoint on an affected model configuration can therefore submit an overlong token_ids list to trigger a worker failure and denial of service. Fixed in 0.29.0.
Severity
SSVC
Exploitation: poc
Automatable: no
Technical Impact: partial
CISA Coordinator · CISA-ADP (v2.0.3)
Decision recorded 2026-09-28 18:21 UTC
CWE
- CWE-400 - Uncontrolled Resource Consumption
Assigner
References
2 references
| URL | Tags |
|---|---|
| https://github.com/vllm-project/vllm/security/adv… | vendor-advisory |
| https://www.vulncheck.com/advisories/vllm-before-… | third-party-advisory |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| vllm-project | vllm |
Affected:
0 , < 0.29.0
(semver)
Unaffected: 0.29.0 (semver) cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:* |
Date Public
2026-09-12 00:00
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-100651",
"options": [
{
"Exploitation": "poc"
},
{
"Automatable": "no"
},
{
"Technical Impact": "partial"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-09-28T18:21:26.899493Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-09-28T18:25:40.222Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"references": [
{
"tags": [
"exploit"
],
"url": "https://github.com/vllm-project/vllm/security/advisories/GHSA-3mqx-f33v-vgp9"
}
],
"title": "CISA ADP Vulnrichment"
}
],
"cna": {
"affected": [
{
"defaultStatus": "unaffected",
"packageURL": "pkg:pypi/vllm",
"product": "vllm",
"vendor": "vllm-project",
"versions": [
{
"lessThan": "0.29.0",
"status": "affected",
"version": "0",
"versionType": "semver"
},
{
"status": "unaffected",
"version": "0.29.0",
"versionType": "semver"
}
]
}
],
"cpeApplicability": [
{
"nodes": [
{
"cpeMatch": [
{
"criteria": "cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:*",
"versionEndExcluding": "0.29.0",
"vulnerable": true
}
],
"negate": false,
"operator": "OR"
}
]
}
],
"credits": [
{
"lang": "en",
"type": "reporter",
"value": "rexpository"
},
{
"lang": "en",
"type": "coordinator",
"value": "jperezdealgaba"
}
],
"datePublic": "2026-09-12T00:00:00.000Z",
"descriptions": [
{
"lang": "en",
"value": "vLLM before 0.29.0 fails to enforce decoder prompt-length validation on the disaggregated serving endpoint /inference/v1/generate. When the request contains a \u0027features\u0027 (multimodal) payload, vllm/entrypoints/serve/disagg/serving.py builds a multimodal EngineInput directly from the caller-supplied token_ids, and GenerateRequest.token_ids (vllm/entrypoints/serve/disagg/protocol.py) is not checked against model_config.max_model_len. For multimodal processors that report skip_prompt_length_check=True (for example Nemotron Parse, Whisper, and FireRedLID), InputProcessor._validate_prompt_len() returns immediately for both encoder and decoder prompts, so an overlong prompt becomes an EngineCoreRequest and reaches the worker input-batch copy into a fixed max_model_len-wide NumPy row. A client able to reach the endpoint on an affected model configuration can therefore submit an overlong token_ids list to trigger a worker failure and denial of service. Fixed in 0.29.0."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "LOW",
"attackRequirements": "NONE",
"attackVector": "NETWORK",
"baseScore": 7.1,
"baseSeverity": "HIGH",
"privilegesRequired": "LOW",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "NONE",
"vectorString": "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",
"version": "4.0",
"vulnAvailabilityImpact": "HIGH",
"vulnConfidentialityImpact": "NONE",
"vulnIntegrityImpact": "NONE"
},
"format": "CVSS"
},
{
"cvssV3_1": {
"attackComplexity": "LOW",
"attackVector": "NETWORK",
"availabilityImpact": "HIGH",
"baseScore": 6.5,
"baseSeverity": "MEDIUM",
"confidentialityImpact": "NONE",
"integrityImpact": "NONE",
"privilegesRequired": "LOW",
"scope": "UNCHANGED",
"userInteraction": "NONE",
"vectorString": "CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H",
"version": "3.1"
},
"format": "CVSS"
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-400",
"description": "Uncontrolled Resource Consumption",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-09-26T13:23:21.141Z",
"orgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"shortName": "VulnCheck"
},
"references": [
{
"name": "GitHub Security Advisory (GHSA-3mqx-f33v-vgp9)",
"tags": [
"vendor-advisory"
],
"url": "https://github.com/vllm-project/vllm/security/advisories/GHSA-3mqx-f33v-vgp9"
},
{
"name": "VulnCheck Advisory: vllm before 0.29.0 Denial of Service via Decoder Prompt Length Bypass",
"tags": [
"third-party-advisory"
],
"url": "https://www.vulncheck.com/advisories/vllm-before-0.29.0-denial-of-service-via-decoder-prompt-length-bypass"
}
],
"title": "vllm before 0.29.0 Denial of Service via Decoder Prompt Length Bypass",
"x_generator": {
"engine": "vulncheck-endgame"
}
}
},
"cveMetadata": {
"assignerOrgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"assignerShortName": "VulnCheck",
"cveId": "CVE-2026-100651",
"datePublished": "2026-09-26T13:23:21.141Z",
"dateReserved": "2026-09-26T02:33:07.899Z",
"dateUpdated": "2026-09-28T18:25:40.222Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-100650 (GCVE-0-2026-100650)
Vulnerability from nvd – Published: 2026-09-26 13:23 – Updated: 2026-10-01 11:06
VLAI
EPSS
VEX
Title
vLLM before 0.29.0 Resource Exhaustion via Unbounded Media Materialization
Summary
vLLM through 0.29.0 fetches and fully materializes remote or inline media before enforcing its documented media controls (the VLLM_MAX_AUDIO_CLIP_FILESIZE_MB compressed-audio size cap, default 25 MB, and the per-modality --limit-mm-per-prompt item limits). Across four ingress paths — the shared media-acquisition layer (HTTPConnection.get_bytes()/async_get_bytes()), the chat completions audio_url/base64 path, the batch speech runner, and the Rust frontend POST /tokenize route — the server reads the entire HTTP response body, base64-decodes the inline payload, or spawns one fetch/decode task per media part, and only then applies the limit (or, on some paths, never applies it). A remote attacker can therefore cause the API server or batch-runner process to allocate memory and consume outbound bandwidth proportional to an attacker-chosen body size or media item count before the request is rejected, resulting in pre-inference memory and bandwidth exhaustion (denial of service). The chat and batch surfaces require an API key when one is configured; the Rust frontend /tokenize route is unauthenticated by design. There is no code execution or data disclosure impact.
Severity
SSVC
Exploitation: poc
Automatable: no
Technical Impact: partial
CISA Coordinator · CISA-ADP (v2.0.3)
Decision recorded 2026-09-30 14:49 UTC
CWE
- CWE-400 - Uncontrolled Resource Consumption
Assigner
References
3 references
| URL | Tags |
|---|---|
| https://github.com/vllm-project/vllm/security/adv… | vendor-advisory |
| https://github.com/vllm-project/vllm/commit/752a3… | patch |
| https://www.vulncheck.com/advisories/vllm-before-… | third-party-advisory |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| vllm-project | vllm |
Affected:
0 , < 0.29.0
(semver)
Unaffected: 0.29.0 (semver) cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:* |
Date Public
2026-09-12 00:00
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-100650",
"options": [
{
"Exploitation": "poc"
},
{
"Automatable": "no"
},
{
"Technical Impact": "partial"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-09-30T14:49:40.322864Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-09-30T14:50:14.406Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"references": [
{
"tags": [
"exploit"
],
"url": "https://github.com/vllm-project/vllm/security/advisories/GHSA-p6g9-7v3x-m8mv"
}
],
"title": "CISA ADP Vulnrichment"
}
],
"cna": {
"affected": [
{
"defaultStatus": "unaffected",
"packageURL": "pkg:pypi/vllm",
"product": "vllm",
"vendor": "vllm-project",
"versions": [
{
"lessThan": "0.29.0",
"status": "affected",
"version": "0",
"versionType": "semver"
},
{
"status": "unaffected",
"version": "0.29.0",
"versionType": "semver"
}
]
}
],
"cpeApplicability": [
{
"nodes": [
{
"cpeMatch": [
{
"criteria": "cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:*",
"versionEndExcluding": "0.29.0",
"vulnerable": true
}
],
"negate": false,
"operator": "OR"
}
]
}
],
"credits": [
{
"lang": "en",
"type": "finder",
"value": "KernelClint"
},
{
"lang": "en",
"type": "finder",
"value": "dhalf"
},
{
"lang": "en",
"type": "coordinator",
"value": "jperezdealgaba"
}
],
"datePublic": "2026-09-12T00:00:00.000Z",
"descriptions": [
{
"lang": "en",
"value": "vLLM through 0.29.0 fetches and fully materializes remote or inline media before enforcing its documented media controls (the VLLM_MAX_AUDIO_CLIP_FILESIZE_MB compressed-audio size cap, default 25 MB, and the per-modality --limit-mm-per-prompt item limits). Across four ingress paths \u2014 the shared media-acquisition layer (HTTPConnection.get_bytes()/async_get_bytes()), the chat completions audio_url/base64 path, the batch speech runner, and the Rust frontend POST /tokenize route \u2014 the server reads the entire HTTP response body, base64-decodes the inline payload, or spawns one fetch/decode task per media part, and only then applies the limit (or, on some paths, never applies it). A remote attacker can therefore cause the API server or batch-runner process to allocate memory and consume outbound bandwidth proportional to an attacker-chosen body size or media item count before the request is rejected, resulting in pre-inference memory and bandwidth exhaustion (denial of service). The chat and batch surfaces require an API key when one is configured; the Rust frontend /tokenize route is unauthenticated by design. There is no code execution or data disclosure impact."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "LOW",
"attackRequirements": "NONE",
"attackVector": "NETWORK",
"baseScore": 7.1,
"baseSeverity": "HIGH",
"privilegesRequired": "LOW",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "NONE",
"vectorString": "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",
"version": "4.0",
"vulnAvailabilityImpact": "HIGH",
"vulnConfidentialityImpact": "NONE",
"vulnIntegrityImpact": "NONE"
},
"format": "CVSS"
},
{
"cvssV3_1": {
"attackComplexity": "LOW",
"attackVector": "NETWORK",
"availabilityImpact": "HIGH",
"baseScore": 6.5,
"baseSeverity": "MEDIUM",
"confidentialityImpact": "NONE",
"integrityImpact": "NONE",
"privilegesRequired": "LOW",
"scope": "UNCHANGED",
"userInteraction": "NONE",
"vectorString": "CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H",
"version": "3.1"
},
"format": "CVSS"
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-400",
"description": "Uncontrolled Resource Consumption",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-10-01T11:06:44.741Z",
"orgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"shortName": "VulnCheck"
},
"references": [
{
"name": "GitHub Security Advisory (GHSA-p6g9-7v3x-m8mv)",
"tags": [
"vendor-advisory"
],
"url": "https://github.com/vllm-project/vllm/security/advisories/GHSA-p6g9-7v3x-m8mv"
},
{
"name": "Patch Commit",
"tags": [
"patch"
],
"url": "https://github.com/vllm-project/vllm/commit/752a3a504485790a2e8491cacbb35c137339ad34"
},
{
"name": "VulnCheck Advisory: vLLM before 0.29.0 Resource Exhaustion via Unbounded Media Materialization",
"tags": [
"third-party-advisory"
],
"url": "https://www.vulncheck.com/advisories/vllm-before-0.29.0-resource-exhaustion-via-unbounded-media-materialization"
}
],
"title": "vLLM before 0.29.0 Resource Exhaustion via Unbounded Media Materialization",
"x_generator": {
"engine": "vulncheck-endgame"
}
}
},
"cveMetadata": {
"assignerOrgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"assignerShortName": "VulnCheck",
"cveId": "CVE-2026-100650",
"datePublished": "2026-09-26T13:23:20.418Z",
"dateReserved": "2026-09-26T02:33:07.898Z",
"dateUpdated": "2026-10-01T11:06:44.741Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-100649 (GCVE-0-2026-100649)
Vulnerability from nvd – Published: 2026-09-26 13:23 – Updated: 2026-09-26 13:23
VLAI
EPSS
VEX
Title
vLLM before 0.29.0 Resource Limit Bypass via Sampler Subclass
Summary
vLLM before 0.29.0 contains a resource-limit bypass vulnerability in PyNvVideoCodec decoder allocation where sampler subclass shadowing allows independent counter increments. Unauthenticated attackers can select different sampler subclasses in video requests to exceed configured decoder limits and exhaust unaccounted GPU memory.
Severity
CWE
- CWE-770 - Allocation of Resources Without Limits or Throttling
Assigner
References
2 references
| URL | Tags |
|---|---|
| https://github.com/vllm-project/vllm/security/adv… | vendor-advisory |
| https://www.vulncheck.com/advisories/vllm-before-… | third-party-advisory |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| vllm-project | vllm |
Affected:
0 , < 0.29.0
(semver)
Unaffected: 0.29.0 (semver) cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:* |
Date Public
2026-09-12 00:00
{
"containers": {
"cna": {
"affected": [
{
"defaultStatus": "unaffected",
"packageURL": "pkg:pypi/vllm",
"product": "vllm",
"vendor": "vllm-project",
"versions": [
{
"lessThan": "0.29.0",
"status": "affected",
"version": "0",
"versionType": "semver"
},
{
"status": "unaffected",
"version": "0.29.0",
"versionType": "semver"
}
]
}
],
"cpeApplicability": [
{
"nodes": [
{
"cpeMatch": [
{
"criteria": "cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:*",
"versionEndExcluding": "0.29.0",
"vulnerable": true
}
],
"negate": false,
"operator": "OR"
}
]
}
],
"credits": [
{
"lang": "en",
"type": "reporter",
"value": "JPengLi"
},
{
"lang": "en",
"type": "coordinator",
"value": "jperezdealgaba"
}
],
"datePublic": "2026-09-12T00:00:00.000Z",
"descriptions": [
{
"lang": "en",
"value": "vLLM before 0.29.0 contains a resource-limit bypass vulnerability in PyNvVideoCodec decoder allocation where sampler subclass shadowing allows independent counter increments. Unauthenticated attackers can select different sampler subclasses in video requests to exceed configured decoder limits and exhaust unaccounted GPU memory."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "HIGH",
"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:H/AT:P/PR:N/UI:N/VC:N/VI:N/VA:L/SC:N/SI:N/SA:N",
"version": "4.0",
"vulnAvailabilityImpact": "LOW",
"vulnConfidentialityImpact": "NONE",
"vulnIntegrityImpact": "NONE"
},
"format": "CVSS"
},
{
"cvssV3_1": {
"attackComplexity": "HIGH",
"attackVector": "NETWORK",
"availabilityImpact": "LOW",
"baseScore": 3.7,
"baseSeverity": "LOW",
"confidentialityImpact": "NONE",
"integrityImpact": "NONE",
"privilegesRequired": "NONE",
"scope": "UNCHANGED",
"userInteraction": "NONE",
"vectorString": "CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:N/A:L",
"version": "3.1"
},
"format": "CVSS"
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-770",
"description": "Allocation of Resources Without Limits or Throttling",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-09-26T13:23:19.687Z",
"orgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"shortName": "VulnCheck"
},
"references": [
{
"name": "GitHub Security Advisory (GHSA-j682-9xp5-rrf3)",
"tags": [
"vendor-advisory"
],
"url": "https://github.com/vllm-project/vllm/security/advisories/GHSA-j682-9xp5-rrf3"
},
{
"name": "VulnCheck Advisory: vLLM before 0.29.0 Resource Limit Bypass via Sampler Subclass",
"tags": [
"third-party-advisory"
],
"url": "https://www.vulncheck.com/advisories/vllm-before-0.29.0-resource-limit-bypass-via-sampler-subclass"
}
],
"title": "vLLM before 0.29.0 Resource Limit Bypass via Sampler Subclass",
"x_generator": {
"engine": "vulncheck-endgame"
}
}
},
"cveMetadata": {
"assignerOrgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"assignerShortName": "VulnCheck",
"cveId": "CVE-2026-100649",
"datePublished": "2026-09-26T13:23:19.687Z",
"dateReserved": "2026-09-26T02:33:07.898Z",
"dateUpdated": "2026-09-26T13:23:19.687Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-100648 (GCVE-0-2026-100648)
Vulnerability from nvd – Published: 2026-09-26 13:23 – Updated: 2026-09-28 16:40
VLAI
EPSS
VEX
Title
vllm before 0.29.0 Uncontrolled Resource Consumption via Audio Decoding
Summary
vllm before 0.29.0 fails to enforce VLLM_MAX_AUDIO_CLIP_FILESIZE_MB limit in multimodal chat audio decoding, allowing unauthenticated clients to bypass file size restrictions. Attackers can submit oversized audio files through chat endpoints to consume excessive memory and CPU resources during decoding.
Severity
5.3 (Medium)
SSVC
Exploitation: none
Automatable: yes
Technical Impact: partial
CISA Coordinator · CISA-ADP (v2.0.3)
Decision recorded 2026-09-28 16:40 UTC
CWE
- CWE-400 - Uncontrolled Resource Consumption
Assigner
References
2 references
| URL | Tags |
|---|---|
| https://github.com/vllm-project/vllm/security/adv… | vendor-advisory |
| https://www.vulncheck.com/advisories/vllm-before-… | third-party-advisory |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| vllm-project | vllm |
Affected:
0 , < 0.29.0
(semver)
Unaffected: 0.29.0 (semver) cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:* |
Date Public
2026-09-12 00:00
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-100648",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "yes"
},
{
"Technical Impact": "partial"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-09-28T16:40:22.871965Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-09-28T16:40:42.006Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"cna": {
"affected": [
{
"defaultStatus": "unaffected",
"packageURL": "pkg:pypi/vllm",
"product": "vllm",
"vendor": "vllm-project",
"versions": [
{
"lessThan": "0.29.0",
"status": "affected",
"version": "0",
"versionType": "semver"
},
{
"status": "unaffected",
"version": "0.29.0",
"versionType": "semver"
}
]
}
],
"cpeApplicability": [
{
"nodes": [
{
"cpeMatch": [
{
"criteria": "cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:*",
"versionEndExcluding": "0.29.0",
"vulnerable": true
}
],
"negate": false,
"operator": "OR"
}
]
}
],
"credits": [
{
"lang": "en",
"type": "reporter",
"value": "santhreal"
},
{
"lang": "en",
"type": "coordinator",
"value": "jperezdealgaba"
}
],
"datePublic": "2026-09-12T00:00:00.000Z",
"descriptions": [
{
"lang": "en",
"value": "vllm before 0.29.0 fails to enforce VLLM_MAX_AUDIO_CLIP_FILESIZE_MB limit in multimodal chat audio decoding, allowing unauthenticated clients to bypass file size restrictions. Attackers can submit oversized audio files through chat endpoints to consume excessive memory and CPU resources during decoding."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "LOW",
"attackRequirements": "NONE",
"attackVector": "NETWORK",
"baseScore": 6.9,
"baseSeverity": "MEDIUM",
"privilegesRequired": "NONE",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "NONE",
"vectorString": "CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:N/VI:N/VA:L/SC:N/SI:N/SA:N",
"version": "4.0",
"vulnAvailabilityImpact": "LOW",
"vulnConfidentialityImpact": "NONE",
"vulnIntegrityImpact": "NONE"
},
"format": "CVSS"
},
{
"cvssV3_1": {
"attackComplexity": "LOW",
"attackVector": "NETWORK",
"availabilityImpact": "LOW",
"baseScore": 5.3,
"baseSeverity": "MEDIUM",
"confidentialityImpact": "NONE",
"integrityImpact": "NONE",
"privilegesRequired": "NONE",
"scope": "UNCHANGED",
"userInteraction": "NONE",
"vectorString": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:L",
"version": "3.1"
},
"format": "CVSS"
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-400",
"description": "Uncontrolled Resource Consumption",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-09-26T13:23:18.965Z",
"orgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"shortName": "VulnCheck"
},
"references": [
{
"name": "GitHub Security Advisory (GHSA-jcq2-4gch-5qhf)",
"tags": [
"vendor-advisory"
],
"url": "https://github.com/vllm-project/vllm/security/advisories/GHSA-jcq2-4gch-5qhf"
},
{
"name": "VulnCheck Advisory: vllm before 0.29.0 Uncontrolled Resource Consumption via Audio Decoding",
"tags": [
"third-party-advisory"
],
"url": "https://www.vulncheck.com/advisories/vllm-before-0.29.0-uncontrolled-resource-consumption-via-audio-decoding"
}
],
"title": "vllm before 0.29.0 Uncontrolled Resource Consumption via Audio Decoding",
"x_generator": {
"engine": "vulncheck-endgame"
}
}
},
"cveMetadata": {
"assignerOrgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"assignerShortName": "VulnCheck",
"cveId": "CVE-2026-100648",
"datePublished": "2026-09-26T13:23:18.965Z",
"dateReserved": "2026-09-26T02:33:07.898Z",
"dateUpdated": "2026-09-28T16:40:42.006Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-100647 (GCVE-0-2026-100647)
Vulnerability from nvd – Published: 2026-09-26 13:23 – Updated: 2026-09-28 18:28
VLAI
EPSS
VEX
Title
vLLM before 0.29.0 CPU Exhaustion via unbounded cache_salt
Summary
vLLM versions before 0.29.0 contain a denial-of-service vulnerability in the cache_salt parameter accepted on OpenAI-compatible and Anthropic API endpoints, which lacks maximum length validation and is processed on the single EngineCore scheduler thread. Unauthenticated attackers can send HTTP requests with multi-hundred-megabyte salt values that trigger expensive pickle serialization and SHA-256 hashing, stalling the scheduler thread and denying service to all concurrent requests.
Severity
5.3 (Medium)
SSVC
Exploitation: none
Automatable: yes
Technical Impact: partial
CISA Coordinator · CISA-ADP (v2.0.3)
Decision recorded 2026-09-28 18:26 UTC
CWE
- CWE-20 - Improper Input Validation
Assigner
References
2 references
| URL | Tags |
|---|---|
| https://github.com/vllm-project/vllm/security/adv… | vendor-advisory |
| https://www.vulncheck.com/advisories/vllm-before-… | third-party-advisory |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| vllm-project | vllm |
Affected:
0 , < 0.29.0
(semver)
Unaffected: 0.29.0 (semver) cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:* |
Date Public
2026-09-12 00:00
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-100647",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "yes"
},
{
"Technical Impact": "partial"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-09-28T18:26:37.598538Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-09-28T18:28:14.772Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"cna": {
"affected": [
{
"defaultStatus": "unaffected",
"packageURL": "pkg:pypi/vllm",
"product": "vllm",
"vendor": "vllm-project",
"versions": [
{
"lessThan": "0.29.0",
"status": "affected",
"version": "0",
"versionType": "semver"
},
{
"status": "unaffected",
"version": "0.29.0",
"versionType": "semver"
}
]
}
],
"cpeApplicability": [
{
"nodes": [
{
"cpeMatch": [
{
"criteria": "cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:*",
"versionEndExcluding": "0.29.0",
"vulnerable": true
}
],
"negate": false,
"operator": "OR"
}
]
}
],
"credits": [
{
"lang": "en",
"type": "reporter",
"value": "wildoranges"
},
{
"lang": "en",
"type": "coordinator",
"value": "jperezdealgaba"
}
],
"datePublic": "2026-09-12T00:00:00.000Z",
"descriptions": [
{
"lang": "en",
"value": "vLLM versions before 0.29.0 contain a denial-of-service vulnerability in the cache_salt parameter accepted on OpenAI-compatible and Anthropic API endpoints, which lacks maximum length validation and is processed on the single EngineCore scheduler thread. Unauthenticated attackers can send HTTP requests with multi-hundred-megabyte salt values that trigger expensive pickle serialization and SHA-256 hashing, stalling the scheduler thread and denying service to all concurrent requests."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "LOW",
"attackRequirements": "NONE",
"attackVector": "NETWORK",
"baseScore": 6.9,
"baseSeverity": "MEDIUM",
"privilegesRequired": "NONE",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "NONE",
"vectorString": "CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:N/VI:N/VA:L/SC:N/SI:N/SA:N",
"version": "4.0",
"vulnAvailabilityImpact": "LOW",
"vulnConfidentialityImpact": "NONE",
"vulnIntegrityImpact": "NONE"
},
"format": "CVSS"
},
{
"cvssV3_1": {
"attackComplexity": "LOW",
"attackVector": "NETWORK",
"availabilityImpact": "LOW",
"baseScore": 5.3,
"baseSeverity": "MEDIUM",
"confidentialityImpact": "NONE",
"integrityImpact": "NONE",
"privilegesRequired": "NONE",
"scope": "UNCHANGED",
"userInteraction": "NONE",
"vectorString": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:L",
"version": "3.1"
},
"format": "CVSS"
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-20",
"description": "Improper Input Validation",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-09-26T13:23:18.240Z",
"orgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"shortName": "VulnCheck"
},
"references": [
{
"name": "GitHub Security Advisory (GHSA-wpww-v874-ph2p)",
"tags": [
"vendor-advisory"
],
"url": "https://github.com/vllm-project/vllm/security/advisories/GHSA-wpww-v874-ph2p"
},
{
"name": "VulnCheck Advisory: vLLM before 0.29.0 CPU Exhaustion via unbounded cache_salt",
"tags": [
"third-party-advisory"
],
"url": "https://www.vulncheck.com/advisories/vllm-before-0.29.0-cpu-exhaustion-via-unbounded-cache-salt"
}
],
"title": "vLLM before 0.29.0 CPU Exhaustion via unbounded cache_salt",
"x_generator": {
"engine": "vulncheck-endgame"
}
}
},
"cveMetadata": {
"assignerOrgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"assignerShortName": "VulnCheck",
"cveId": "CVE-2026-100647",
"datePublished": "2026-09-26T13:23:18.240Z",
"dateReserved": "2026-09-26T02:33:07.898Z",
"dateUpdated": "2026-09-28T18:28:14.772Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-94627 (GCVE-0-2026-94627)
Vulnerability from nvd – Published: 2026-09-21 22:04 – Updated: 2026-09-22 12:51
VLAI
EPSS
VEX
Title
vLLM through 0.29.0 GPU KV Cache Leak via Mooncake Transfer ID Collision
Summary
vLLM Mooncake connector through 0.29.0 fails to properly manage GPU KV cache block ownership when concurrent child requests share a single transfer ID in prefill/decode disaggregated deployments. Attackers can trigger GPU memory exhaustion by submitting completion requests with multiple prompts, causing orphaned KV cache blocks to accumulate until process restart and eventually preventing legitimate requests from executing.
Severity
SSVC
Exploitation: none
Automatable: yes
Technical Impact: partial
CISA Coordinator · CISA-ADP (v2.0.3)
Decision recorded 2026-09-22 12:50 UTC
CWE
- CWE-401 - Missing Release of Memory after Effective Lifetime
Assigner
References
4 references
| URL | Tags |
|---|---|
| https://github.com/vllm-project/vllm/pull/49796 | issue-trackingpatch |
| https://github.com/vllm-project/vllm/blob/v0.29.0… | technical-description |
| https://github.com/vllm-project/vllm | product |
| https://www.vulncheck.com/advisories/vllm-through… | third-party-advisory |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| vllm-project | vllm |
Affected:
0 , ≤ 0.29.0
(semver)
cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:* |
Date Public
2026-07-25 00:00
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-94627",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "yes"
},
{
"Technical Impact": "partial"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-09-22T12:50:58.697049Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-09-22T12:51:23.062Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"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": "finder",
"value": "Mingkai Yu"
},
{
"lang": "en",
"type": "finder",
"value": "Jiapeng Li"
},
{
"lang": "en",
"type": "finder",
"value": "Jiajia Liu"
}
],
"datePublic": "2026-07-25T00:00:00.000Z",
"descriptions": [
{
"lang": "en",
"value": "vLLM Mooncake connector through 0.29.0 fails to properly manage GPU KV cache block ownership when concurrent child requests share a single transfer ID in prefill/decode disaggregated deployments. Attackers can trigger GPU memory exhaustion by submitting completion requests with multiple prompts, causing orphaned KV cache blocks to accumulate until process restart and eventually preventing legitimate requests from executing."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "LOW",
"attackRequirements": "NONE",
"attackVector": "NETWORK",
"baseScore": 8.7,
"baseSeverity": "HIGH",
"privilegesRequired": "NONE",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "NONE",
"vectorString": "CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:N/VI:N/VA:H/SC:N/SI:N/SA:N",
"version": "4.0",
"vulnAvailabilityImpact": "HIGH",
"vulnConfidentialityImpact": "NONE",
"vulnIntegrityImpact": "NONE"
},
"format": "CVSS"
},
{
"cvssV3_1": {
"attackComplexity": "LOW",
"attackVector": "NETWORK",
"availabilityImpact": "HIGH",
"baseScore": 7.5,
"baseSeverity": "HIGH",
"confidentialityImpact": "NONE",
"integrityImpact": "NONE",
"privilegesRequired": "NONE",
"scope": "UNCHANGED",
"userInteraction": "NONE",
"vectorString": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H",
"version": "3.1"
},
"format": "CVSS"
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-401",
"description": "Missing Release of Memory after Effective Lifetime",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-09-21T22:04:16.559Z",
"orgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"shortName": "VulnCheck"
},
"references": [
{
"name": "Pull Request #49796",
"tags": [
"issue-tracking",
"patch"
],
"url": "https://github.com/vllm-project/vllm/pull/49796"
},
{
"tags": [
"technical-description"
],
"url": "https://github.com/vllm-project/vllm/blob/v0.29.0/vllm/distributed/kv_transfer/kv_connector/v1/mooncake/mooncake_connector.py#L1978-L1989"
},
{
"tags": [
"product"
],
"url": "https://github.com/vllm-project/vllm"
},
{
"name": "VulnCheck Advisory: vLLM through 0.29.0 GPU KV Cache Leak via Mooncake Transfer ID Collision",
"tags": [
"third-party-advisory"
],
"url": "https://www.vulncheck.com/advisories/vllm-through-0.29.0-gpu-kv-cache-leak-via-mooncake-transfer-id-collision"
}
],
"title": "vLLM through 0.29.0 GPU KV Cache Leak via Mooncake Transfer ID Collision",
"x_generator": {
"engine": "vulncheck-endgame"
}
}
},
"cveMetadata": {
"assignerOrgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"assignerShortName": "VulnCheck",
"cveId": "CVE-2026-94627",
"datePublished": "2026-09-21T22:04:16.559Z",
"dateReserved": "2026-09-21T21:42:30.080Z",
"dateUpdated": "2026-09-22T12:51:23.062Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-94626 (GCVE-0-2026-94626)
Vulnerability from nvd – Published: 2026-09-21 22:04 – Updated: 2026-09-24 22:54
VLAI
EPSS
VEX
Title
vLLM through 0.29.0 Memory Exhaustion via Unvalidated NIXL tp_size
Summary
vLLM through 0.29.0 fails to validate the tp_size parameter in kv_transfer_params on OpenAI-compatible completion endpoints, allowing attackers to allocate unbounded memory. Attackers can supply arbitrary tp_size values in prefill/decode disaggregated deployments to exhaust memory and trigger kernel OOM-kill of the decode worker process.
Severity
SSVC
Exploitation: none
Automatable: yes
Technical Impact: partial
CISA Coordinator · CISA-ADP (v2.0.3)
Decision recorded 2026-09-24 22:54 UTC
CWE
- CWE-789 - Memory Allocation with Excessive Size Value
Assigner
References
5 references
| URL | Tags |
|---|---|
| https://github.com/vllm-project/vllm/pull/51137 | issue-trackingpatch |
| 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://github.com/vllm-project/vllm | product |
| https://www.vulncheck.com/advisories/vllm-through… | third-party-advisory |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| vllm-project | vllm |
Affected:
0 , ≤ 0.29.0
(semver)
cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:* |
Date Public
2026-08-05 00:00
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-94626",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "yes"
},
{
"Technical Impact": "partial"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-09-24T22:54:32.814414Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-09-24T22:54:41.241Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"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": "finder",
"value": "Mingkai Yu"
},
{
"lang": "en",
"type": "finder",
"value": "Jiapeng Li"
},
{
"lang": "en",
"type": "finder",
"value": "Jiajia Liu"
}
],
"datePublic": "2026-08-05T00:00:00.000Z",
"descriptions": [
{
"lang": "en",
"value": "vLLM through 0.29.0 fails to validate the tp_size parameter in kv_transfer_params on OpenAI-compatible completion endpoints, allowing attackers to allocate unbounded memory. Attackers can supply arbitrary tp_size values in prefill/decode disaggregated deployments to exhaust memory and trigger kernel OOM-kill of the decode worker process."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "LOW",
"attackRequirements": "NONE",
"attackVector": "NETWORK",
"baseScore": 8.7,
"baseSeverity": "HIGH",
"privilegesRequired": "NONE",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "NONE",
"vectorString": "CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:N/VI:N/VA:H/SC:N/SI:N/SA:N",
"version": "4.0",
"vulnAvailabilityImpact": "HIGH",
"vulnConfidentialityImpact": "NONE",
"vulnIntegrityImpact": "NONE"
},
"format": "CVSS"
},
{
"cvssV3_1": {
"attackComplexity": "LOW",
"attackVector": "NETWORK",
"availabilityImpact": "HIGH",
"baseScore": 7.5,
"baseSeverity": "HIGH",
"confidentialityImpact": "NONE",
"integrityImpact": "NONE",
"privilegesRequired": "NONE",
"scope": "UNCHANGED",
"userInteraction": "NONE",
"vectorString": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H",
"version": "3.1"
},
"format": "CVSS"
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-789",
"description": "Memory Allocation with Excessive Size Value",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-09-21T22:04:15.352Z",
"orgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"shortName": "VulnCheck"
},
"references": [
{
"name": "Pull Request #51137",
"tags": [
"issue-tracking",
"patch"
],
"url": "https://github.com/vllm-project/vllm/pull/51137"
},
{
"tags": [
"technical-description"
],
"url": "https://github.com/vllm-project/vllm/blob/v0.29.0/vllm/distributed/kv_transfer/kv_connector/utils.py#L569-L573"
},
{
"tags": [
"technical-description"
],
"url": "https://github.com/vllm-project/vllm/blob/v0.29.0/vllm/distributed/kv_transfer/kv_connector/v1/nixl/metadata.py#L277"
},
{
"tags": [
"product"
],
"url": "https://github.com/vllm-project/vllm"
},
{
"name": "VulnCheck Advisory: vLLM through 0.29.0 Memory Exhaustion via Unvalidated NIXL tp_size",
"tags": [
"third-party-advisory"
],
"url": "https://www.vulncheck.com/advisories/vllm-through-0.29.0-memory-exhaustion-via-unvalidated-nixl-tp-size"
}
],
"title": "vLLM through 0.29.0 Memory Exhaustion via Unvalidated NIXL tp_size",
"x_generator": {
"engine": "vulncheck-endgame"
}
}
},
"cveMetadata": {
"assignerOrgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"assignerShortName": "VulnCheck",
"cveId": "CVE-2026-94626",
"datePublished": "2026-09-21T22:04:15.352Z",
"dateReserved": "2026-09-21T21:42:28.781Z",
"dateUpdated": "2026-09-24T22:54:41.241Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-94625 (GCVE-0-2026-94625)
Vulnerability from nvd – Published: 2026-09-21 22:04 – Updated: 2026-09-23 15:09
VLAI
EPSS
VEX
Title
vLLM through 0.29.0 Resource Exhaustion via Ownerless Mooncake Transfer Placeholders
Summary
vLLM through 0.29.0 contains a resource exhaustion vulnerability in MooncakeConnector where rejected prefill requests create ownerless transfer placeholders that are never reclaimed. Attackers can send rejected requests to exhaust sender task pools, causing valid requests to be delayed by up to 480 seconds while health checks continue returning success.
Severity
5.3 (Medium)
SSVC
Exploitation: none
Automatable: yes
Technical Impact: partial
CISA Coordinator · CISA-ADP (v2.0.3)
Decision recorded 2026-09-23 15:09 UTC
CWE
- CWE-772 - Missing Release of Resource after Effective Lifetime
Assigner
References
4 references
| URL | Tags |
|---|---|
| https://github.com/vllm-project/vllm/pull/51236 | issue-trackingpatch |
| https://github.com/vllm-project/vllm/blob/v0.29.0… | technical-description |
| https://github.com/vllm-project/vllm | product |
| https://www.vulncheck.com/advisories/vllm-through… | third-party-advisory |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| vllm-project | vllm |
Affected:
0 , ≤ 0.29.0
(semver)
cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:* |
Date Public
2026-08-06 00:00
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-94625",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "yes"
},
{
"Technical Impact": "partial"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-09-23T15:09:07.861181Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-09-23T15:09:16.926Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"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": "finder",
"value": "Mingkai Yu"
},
{
"lang": "en",
"type": "finder",
"value": "Jiapeng Li"
},
{
"lang": "en",
"type": "finder",
"value": "Jiajia Liu"
}
],
"datePublic": "2026-08-06T00:00:00.000Z",
"descriptions": [
{
"lang": "en",
"value": "vLLM through 0.29.0 contains a resource exhaustion vulnerability in MooncakeConnector where rejected prefill requests create ownerless transfer placeholders that are never reclaimed. Attackers can send rejected requests to exhaust sender task pools, causing valid requests to be delayed by up to 480 seconds while health checks continue returning success."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "LOW",
"attackRequirements": "NONE",
"attackVector": "NETWORK",
"baseScore": 6.9,
"baseSeverity": "MEDIUM",
"privilegesRequired": "NONE",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "NONE",
"vectorString": "CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:N/VI:N/VA:L/SC:N/SI:N/SA:N",
"version": "4.0",
"vulnAvailabilityImpact": "LOW",
"vulnConfidentialityImpact": "NONE",
"vulnIntegrityImpact": "NONE"
},
"format": "CVSS"
},
{
"cvssV3_1": {
"attackComplexity": "LOW",
"attackVector": "NETWORK",
"availabilityImpact": "LOW",
"baseScore": 5.3,
"baseSeverity": "MEDIUM",
"confidentialityImpact": "NONE",
"integrityImpact": "NONE",
"privilegesRequired": "NONE",
"scope": "UNCHANGED",
"userInteraction": "NONE",
"vectorString": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:L",
"version": "3.1"
},
"format": "CVSS"
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-772",
"description": "Missing Release of Resource after Effective Lifetime",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-09-21T22:04:14.206Z",
"orgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"shortName": "VulnCheck"
},
"references": [
{
"name": "Pull Request #51236",
"tags": [
"issue-tracking",
"patch"
],
"url": "https://github.com/vllm-project/vllm/pull/51236"
},
{
"tags": [
"technical-description"
],
"url": "https://github.com/vllm-project/vllm/blob/v0.29.0/vllm/distributed/kv_transfer/kv_connector/v1/mooncake/mooncake_connector.py#L1234-L1242"
},
{
"tags": [
"product"
],
"url": "https://github.com/vllm-project/vllm"
},
{
"name": "VulnCheck Advisory: vLLM through 0.29.0 Resource Exhaustion via Ownerless Mooncake Transfer Placeholders",
"tags": [
"third-party-advisory"
],
"url": "https://www.vulncheck.com/advisories/vllm-through-0.29.0-resource-exhaustion-via-ownerless-mooncake-transfer-placeholders"
}
],
"title": "vLLM through 0.29.0 Resource Exhaustion via Ownerless Mooncake Transfer Placeholders",
"x_generator": {
"engine": "vulncheck-endgame"
}
}
},
"cveMetadata": {
"assignerOrgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"assignerShortName": "VulnCheck",
"cveId": "CVE-2026-94625",
"datePublished": "2026-09-21T22:04:14.206Z",
"dateReserved": "2026-09-21T21:42:27.917Z",
"dateUpdated": "2026-09-23T15:09:16.926Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-94624 (GCVE-0-2026-94624)
Vulnerability from nvd – Published: 2026-09-21 22:04 – Updated: 2026-09-22 13:26
VLAI
EPSS
VEX
Title
vLLM through 0.29.0 Denial of Service via Unbounded P2P KV Offloading Sessions
Summary
vLLM through 0.29.0 contains a denial of service vulnerability in P2P KV offloading when OffloadingConnector is configured with TieringOffloadingSpec and a peer-to-peer secondary tier. Attackers can supply arbitrary remote host and port values in kv_transfer_params to create unreachable peer sessions that retain ZeroMQ sockets until the context quota is exhausted, causing an uncaught ZMQError that crashes EngineCore and stops all inference.
Severity
SSVC
Exploitation: none
Automatable: yes
Technical Impact: partial
CISA Coordinator · CISA-ADP (v2.0.3)
Decision recorded 2026-09-22 13:26 UTC
CWE
- CWE-770 - Allocation of Resources Without Limits or Throttling
Assigner
References
5 references
| URL | Tags |
|---|---|
| https://github.com/vllm-project/vllm/pull/51504 | issue-trackingpatch |
| 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://github.com/vllm-project/vllm | product |
| https://www.vulncheck.com/advisories/vllm-through… | third-party-advisory |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| vllm-project | vllm |
Affected:
0 , ≤ 0.29.0
(semver)
cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:* |
Date Public
2026-08-08 00:00
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-94624",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "yes"
},
{
"Technical Impact": "partial"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-09-22T13:26:32.508111Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-09-22T13:26:57.950Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"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": "finder",
"value": "Mingkai Yu"
},
{
"lang": "en",
"type": "finder",
"value": "Jiapeng Li"
},
{
"lang": "en",
"type": "finder",
"value": "Jiajia Liu"
}
],
"datePublic": "2026-08-08T00:00:00.000Z",
"descriptions": [
{
"lang": "en",
"value": "vLLM through 0.29.0 contains a denial of service vulnerability in P2P KV offloading when OffloadingConnector is configured with TieringOffloadingSpec and a peer-to-peer secondary tier. Attackers can supply arbitrary remote host and port values in kv_transfer_params to create unreachable peer sessions that retain ZeroMQ sockets until the context quota is exhausted, causing an uncaught ZMQError that crashes EngineCore and stops all inference."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "LOW",
"attackRequirements": "NONE",
"attackVector": "NETWORK",
"baseScore": 8.7,
"baseSeverity": "HIGH",
"privilegesRequired": "NONE",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "NONE",
"vectorString": "CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:N/VI:N/VA:H/SC:N/SI:N/SA:N",
"version": "4.0",
"vulnAvailabilityImpact": "HIGH",
"vulnConfidentialityImpact": "NONE",
"vulnIntegrityImpact": "NONE"
},
"format": "CVSS"
},
{
"cvssV3_1": {
"attackComplexity": "LOW",
"attackVector": "NETWORK",
"availabilityImpact": "HIGH",
"baseScore": 7.5,
"baseSeverity": "HIGH",
"confidentialityImpact": "NONE",
"integrityImpact": "NONE",
"privilegesRequired": "NONE",
"scope": "UNCHANGED",
"userInteraction": "NONE",
"vectorString": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H",
"version": "3.1"
},
"format": "CVSS"
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-770",
"description": "Allocation of Resources Without Limits or Throttling",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-09-21T22:04:13.172Z",
"orgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"shortName": "VulnCheck"
},
"references": [
{
"name": "Pull Request #51504",
"tags": [
"issue-tracking",
"patch"
],
"url": "https://github.com/vllm-project/vllm/pull/51504"
},
{
"tags": [
"technical-description"
],
"url": "https://github.com/vllm-project/vllm/blob/v0.29.0/vllm/v1/kv_offload/tiering/p2p/control/zmq.py#L240-L260"
},
{
"tags": [
"technical-description"
],
"url": "https://github.com/vllm-project/vllm/blob/v0.29.0/vllm/v1/kv_offload/tiering/p2p/manager.py#L667-L677"
},
{
"tags": [
"product"
],
"url": "https://github.com/vllm-project/vllm"
},
{
"name": "VulnCheck Advisory: vLLM through 0.29.0 Denial of Service via Unbounded P2P KV Offloading Sessions",
"tags": [
"third-party-advisory"
],
"url": "https://www.vulncheck.com/advisories/vllm-through-0.29.0-denial-of-service-via-unbounded-p2p-kv-offloading-sessions"
}
],
"title": "vLLM through 0.29.0 Denial of Service via Unbounded P2P KV Offloading Sessions",
"x_generator": {
"engine": "vulncheck-endgame"
}
}
},
"cveMetadata": {
"assignerOrgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"assignerShortName": "VulnCheck",
"cveId": "CVE-2026-94624",
"datePublished": "2026-09-21T22:04:13.172Z",
"dateReserved": "2026-09-21T21:42:26.965Z",
"dateUpdated": "2026-09-22T13:26:57.950Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-94623 (GCVE-0-2026-94623)
Vulnerability from nvd – Published: 2026-09-21 22:04 – Updated: 2026-09-22 17:44
VLAI
EPSS
VEX
Title
vLLM through 0.29.0 Denial of Service via NIXL Multi-Prompt Assertion Failure
Summary
vLLM through 0.29.0 contains a denial of service vulnerability in the NIXL connector's prefix caching implementation that fails to properly validate block counts across multi-prompt completion requests in prefill/decode disaggregated deployments. Attackers can trigger an assertion failure in NixlBaseConnectorWorker._apply_prefix_caching by submitting completion requests with multiple prompts of varying lengths, causing the decode worker to terminate and become unavailable until restarted.
Severity
SSVC
Exploitation: none
Automatable: yes
Technical Impact: partial
CISA Coordinator · CISA-ADP (v2.0.3)
Decision recorded 2026-09-22 17:19 UTC
CWE
- CWE-617 - Reachable Assertion
Assigner
References
4 references
| URL | Tags |
|---|---|
| https://github.com/vllm-project/vllm/pull/51505 | issue-trackingpatch |
| https://github.com/vllm-project/vllm/blob/v0.29.0… | technical-description |
| https://github.com/vllm-project/vllm | product |
| https://www.vulncheck.com/advisories/vllm-through… | third-party-advisory |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| vllm-project | vllm |
Affected:
0 , ≤ 0.29.0
(semver)
cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:* |
Date Public
2026-08-08 00:00
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-94623",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "yes"
},
{
"Technical Impact": "partial"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-09-22T17:19:44.934826Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-09-22T17:44:56.345Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"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": "finder",
"value": "Mingkai Yu"
},
{
"lang": "en",
"type": "finder",
"value": "Jiapeng Li"
},
{
"lang": "en",
"type": "finder",
"value": "Jiajia Liu"
}
],
"datePublic": "2026-08-08T00:00:00.000Z",
"descriptions": [
{
"lang": "en",
"value": "vLLM through 0.29.0 contains a denial of service vulnerability in the NIXL connector\u0027s prefix caching implementation that fails to properly validate block counts across multi-prompt completion requests in prefill/decode disaggregated deployments. Attackers can trigger an assertion failure in NixlBaseConnectorWorker._apply_prefix_caching by submitting completion requests with multiple prompts of varying lengths, causing the decode worker to terminate and become unavailable until restarted."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "LOW",
"attackRequirements": "NONE",
"attackVector": "NETWORK",
"baseScore": 8.7,
"baseSeverity": "HIGH",
"privilegesRequired": "NONE",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "NONE",
"vectorString": "CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:N/VI:N/VA:H/SC:N/SI:N/SA:N",
"version": "4.0",
"vulnAvailabilityImpact": "HIGH",
"vulnConfidentialityImpact": "NONE",
"vulnIntegrityImpact": "NONE"
},
"format": "CVSS"
},
{
"cvssV3_1": {
"attackComplexity": "LOW",
"attackVector": "NETWORK",
"availabilityImpact": "HIGH",
"baseScore": 7.5,
"baseSeverity": "HIGH",
"confidentialityImpact": "NONE",
"integrityImpact": "NONE",
"privilegesRequired": "NONE",
"scope": "UNCHANGED",
"userInteraction": "NONE",
"vectorString": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H",
"version": "3.1"
},
"format": "CVSS"
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-617",
"description": "Reachable Assertion",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-09-21T22:04:12.106Z",
"orgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"shortName": "VulnCheck"
},
"references": [
{
"name": "Pull Request #51505",
"tags": [
"issue-tracking",
"patch"
],
"url": "https://github.com/vllm-project/vllm/pull/51505"
},
{
"tags": [
"technical-description"
],
"url": "https://github.com/vllm-project/vllm/blob/v0.29.0/vllm/distributed/kv_transfer/kv_connector/v1/nixl/base_worker.py#L2642-L2646"
},
{
"tags": [
"product"
],
"url": "https://github.com/vllm-project/vllm"
},
{
"name": "VulnCheck Advisory: vLLM through 0.29.0 Denial of Service via NIXL Multi-Prompt Assertion Failure",
"tags": [
"third-party-advisory"
],
"url": "https://www.vulncheck.com/advisories/vllm-through-0.29.0-denial-of-service-via-nixl-multi-prompt-assertion-failure"
}
],
"title": "vLLM through 0.29.0 Denial of Service via NIXL Multi-Prompt Assertion Failure",
"x_generator": {
"engine": "vulncheck-endgame"
}
}
},
"cveMetadata": {
"assignerOrgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"assignerShortName": "VulnCheck",
"cveId": "CVE-2026-94623",
"datePublished": "2026-09-21T22:04:12.106Z",
"dateReserved": "2026-09-21T21:42:25.636Z",
"dateUpdated": "2026-09-22T17:44:56.345Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-94622 (GCVE-0-2026-94622)
Vulnerability from nvd – Published: 2026-09-21 22:04 – Updated: 2026-09-22 13:01
VLAI
EPSS
VEX
Title
vLLM through 0.29.0 Denial of Service via Incomplete NIXL KV Transfer Metadata
Summary
vLLM versions through 0.29.0 contain a denial of service vulnerability in the NIXL connector's metadata handling for prefill/decode disaggregated deployments. Attackers can send requests with incomplete kv_transfer_params dictionary entries to trigger an uncaught KeyError in EngineCore scheduling, causing the decode engine to terminate and making all routed requests fail until manual restart.
Severity
SSVC
Exploitation: none
Automatable: yes
Technical Impact: partial
CISA Coordinator · CISA-ADP (v2.0.3)
Decision recorded 2026-09-22 12:59 UTC
CWE
- CWE-248 - Uncaught Exception
Assigner
References
4 references
| URL | Tags |
|---|---|
| https://github.com/vllm-project/vllm/pull/54807 | issue-trackingpatch |
| https://github.com/vllm-project/vllm/blob/v0.29.0… | technical-description |
| https://github.com/vllm-project/vllm | product |
| https://www.vulncheck.com/advisories/vllm-through… | third-party-advisory |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| vllm-project | vllm |
Affected:
0 , ≤ 0.29.0
(semver)
cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:* |
Date Public
2026-09-01 00:00
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-94622",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "yes"
},
{
"Technical Impact": "partial"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-09-22T12:59:50.196633Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-09-22T13:01:33.609Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"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": "finder",
"value": "Mingkai Yu"
},
{
"lang": "en",
"type": "finder",
"value": "Jiapeng Li"
},
{
"lang": "en",
"type": "finder",
"value": "Jiajia Liu"
}
],
"datePublic": "2026-09-01T00:00:00.000Z",
"descriptions": [
{
"lang": "en",
"value": "vLLM versions through 0.29.0 contain a denial of service vulnerability in the NIXL connector\u0027s metadata handling for prefill/decode disaggregated deployments. Attackers can send requests with incomplete kv_transfer_params dictionary entries to trigger an uncaught KeyError in EngineCore scheduling, causing the decode engine to terminate and making all routed requests fail until manual restart."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "LOW",
"attackRequirements": "NONE",
"attackVector": "NETWORK",
"baseScore": 8.7,
"baseSeverity": "HIGH",
"privilegesRequired": "NONE",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "NONE",
"vectorString": "CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:N/VI:N/VA:H/SC:N/SI:N/SA:N",
"version": "4.0",
"vulnAvailabilityImpact": "HIGH",
"vulnConfidentialityImpact": "NONE",
"vulnIntegrityImpact": "NONE"
},
"format": "CVSS"
},
{
"cvssV3_1": {
"attackComplexity": "LOW",
"attackVector": "NETWORK",
"availabilityImpact": "HIGH",
"baseScore": 7.5,
"baseSeverity": "HIGH",
"confidentialityImpact": "NONE",
"integrityImpact": "NONE",
"privilegesRequired": "NONE",
"scope": "UNCHANGED",
"userInteraction": "NONE",
"vectorString": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H",
"version": "3.1"
},
"format": "CVSS"
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-248",
"description": "Uncaught Exception",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-09-21T22:04:10.761Z",
"orgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"shortName": "VulnCheck"
},
"references": [
{
"name": "Pull Request #54807",
"tags": [
"issue-tracking",
"patch"
],
"url": "https://github.com/vllm-project/vllm/pull/54807"
},
{
"tags": [
"technical-description"
],
"url": "https://github.com/vllm-project/vllm/blob/v0.29.0/vllm/distributed/kv_transfer/kv_connector/v1/nixl/metadata.py#L294-L312"
},
{
"tags": [
"product"
],
"url": "https://github.com/vllm-project/vllm"
},
{
"name": "VulnCheck Advisory: vLLM through 0.29.0 Denial of Service via Incomplete NIXL KV Transfer Metadata",
"tags": [
"third-party-advisory"
],
"url": "https://www.vulncheck.com/advisories/vllm-through-0.29.0-denial-of-service-via-incomplete-nixl-kv-transfer-metadata"
}
],
"title": "vLLM through 0.29.0 Denial of Service via Incomplete NIXL KV Transfer Metadata",
"x_generator": {
"engine": "vulncheck-endgame"
}
}
},
"cveMetadata": {
"assignerOrgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"assignerShortName": "VulnCheck",
"cveId": "CVE-2026-94622",
"datePublished": "2026-09-21T22:04:10.761Z",
"dateReserved": "2026-09-21T21:42:24.373Z",
"dateUpdated": "2026-09-22T13:01:33.609Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-103241 (GCVE-0-2026-103241)
Vulnerability from cvelistv5 – Published: 2026-09-30 16:45 – Updated: 2026-09-30 16:45 X_Open Source
VLAI
EPSS
VEX
Title
vllm-project vLLM Gemma4UnifiedParser gemma4.rs denial of service
Summary
A flaw has been found in vllm-project vLLM up to 0.26.0. This vulnerability affects unknown code of the file rust/src/parser/src/unified/gemma4.rs of the component Gemma4UnifiedParser. Executing a manipulation can lead to denial of service. The attack may be launched remotely. The exploit has been published and may be used. Upgrading to version 0.29.1rc0 is able to resolve this issue. This patch is called 3439bad37e68ba9755a46f4f6b44a4aeaf1f60a9. Upgrading the affected component is advised.
Severity
CWE
- CWE-404 - Denial of Service
Assigner
References
10 references
| URL | Tags |
|---|---|
| https://vuldb.com/vuln/411965 | vdb-entry |
| https://vuldb.com/vuln/411965/cti | signaturepermissions-required |
| https://vuldb.com/cve/CVE-2026-103241 | third-party-advisory |
| https://vuldb.com/submit/956250 | third-party-advisory |
| https://github.com/vllm-project/vllm/issues/50927 | issue-tracking |
| https://github.com/vllm-project/vllm/pull/54303 | issue-trackingpatch |
| https://gist.github.com/Yunzez/8e98d656aa667095b5… | exploit |
| https://github.com/vllm-project/vllm/commit/3439b… | patch |
| https://github.com/vllm-project/vllm/releases/tag… | patch |
| https://github.com/vllm-project/vllm/ | product |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| vllm-project | vLLM |
Affected:
0.1
Affected: 0.2 Affected: 0.3 Affected: 0.4 Affected: 0.5 Affected: 0.6 Affected: 0.7 Affected: 0.8 Affected: 0.9 Affected: 0.10 Affected: 0.11 Affected: 0.12 Affected: 0.13 Affected: 0.14 Affected: 0.15 Affected: 0.16 Affected: 0.17 Affected: 0.18 Affected: 0.19 Affected: 0.20 Affected: 0.21 Affected: 0.22 Affected: 0.23 Affected: 0.24 Affected: 0.25 Affected: 0.26.0 Unaffected: 0.29.1rc0 cpe:2.3:a:vllm-project:vllm:*:*:*:*:*:*:*:* |
{
"containers": {
"cna": {
"affected": [
{
"cpes": [
"cpe:2.3:a:vllm-project:vllm:*:*:*:*:*:*:*:*"
],
"modules": [
"Gemma4UnifiedParser"
],
"product": "vLLM",
"vendor": "vllm-project",
"versions": [
{
"status": "affected",
"version": "0.1"
},
{
"status": "affected",
"version": "0.2"
},
{
"status": "affected",
"version": "0.3"
},
{
"status": "affected",
"version": "0.4"
},
{
"status": "affected",
"version": "0.5"
},
{
"status": "affected",
"version": "0.6"
},
{
"status": "affected",
"version": "0.7"
},
{
"status": "affected",
"version": "0.8"
},
{
"status": "affected",
"version": "0.9"
},
{
"status": "affected",
"version": "0.10"
},
{
"status": "affected",
"version": "0.11"
},
{
"status": "affected",
"version": "0.12"
},
{
"status": "affected",
"version": "0.13"
},
{
"status": "affected",
"version": "0.14"
},
{
"status": "affected",
"version": "0.15"
},
{
"status": "affected",
"version": "0.16"
},
{
"status": "affected",
"version": "0.17"
},
{
"status": "affected",
"version": "0.18"
},
{
"status": "affected",
"version": "0.19"
},
{
"status": "affected",
"version": "0.20"
},
{
"status": "affected",
"version": "0.21"
},
{
"status": "affected",
"version": "0.22"
},
{
"status": "affected",
"version": "0.23"
},
{
"status": "affected",
"version": "0.24"
},
{
"status": "affected",
"version": "0.25"
},
{
"status": "affected",
"version": "0.26.0"
},
{
"status": "unaffected",
"version": "0.29.1rc0"
}
]
}
],
"credits": [
{
"lang": "en",
"type": "reporter",
"value": "Zyz3366 (VulDB User)"
}
],
"descriptions": [
{
"lang": "en",
"value": "A flaw has been found in vllm-project vLLM up to 0.26.0. This vulnerability affects unknown code of the file rust/src/parser/src/unified/gemma4.rs of the component Gemma4UnifiedParser. Executing a manipulation can lead to denial of service. The attack may be launched remotely. The exploit has been published and may be used. Upgrading to version 0.29.1rc0 is able to resolve this issue. This patch is called 3439bad37e68ba9755a46f4f6b44a4aeaf1f60a9. Upgrading the affected component is advised."
}
],
"metrics": [
{
"cvssV4_0": {
"baseScore": 6.9,
"baseSeverity": "MEDIUM",
"vectorString": "CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:N/VI:N/VA:L/SC:N/SI:N/SA:N/E:P",
"version": "4.0"
}
},
{
"cvssV3_1": {
"baseScore": 5.3,
"baseSeverity": "MEDIUM",
"vectorString": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:L/E:P/RL:O/RC:C",
"version": "3.1"
}
},
{
"cvssV3_0": {
"baseScore": 5.3,
"baseSeverity": "MEDIUM",
"vectorString": "CVSS:3.0/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:L/E:P/RL:O/RC:C",
"version": "3.0"
}
},
{
"cvssV2_0": {
"baseScore": 5,
"vectorString": "AV:N/AC:L/Au:N/C:N/I:N/A:P/E:POC/RL:OF/RC:C",
"version": "2.0"
}
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-404",
"description": "Denial of Service",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-09-30T16:45:13.581Z",
"orgId": "1af790b2-7ee1-4545-860a-a788eba489b5",
"shortName": "VulDB"
},
"references": [
{
"name": "VDB-411965 | vllm-project vLLM Gemma4UnifiedParser gemma4.rs denial of service",
"tags": [
"vdb-entry"
],
"url": "https://vuldb.com/vuln/411965"
},
{
"name": "VDB-411965 | CTI Indicators (IOB, IOC, TTP, IOA)",
"tags": [
"signature",
"permissions-required"
],
"url": "https://vuldb.com/vuln/411965/cti"
},
{
"name": "CVE-2026-103241 | CVE Analysis and Report",
"tags": [
"third-party-advisory"
],
"url": "https://vuldb.com/cve/CVE-2026-103241"
},
{
"name": "Submit #956250 | vLLM Project vLLM v0.26.0 Denial of Service",
"tags": [
"third-party-advisory"
],
"url": "https://vuldb.com/submit/956250"
},
{
"tags": [
"issue-tracking"
],
"url": "https://github.com/vllm-project/vllm/issues/50927"
},
{
"tags": [
"issue-tracking",
"patch"
],
"url": "https://github.com/vllm-project/vllm/pull/54303"
},
{
"tags": [
"exploit"
],
"url": "https://gist.github.com/Yunzez/8e98d656aa667095b513161eb056d28e"
},
{
"tags": [
"patch"
],
"url": "https://github.com/vllm-project/vllm/commit/3439bad37e68ba9755a46f4f6b44a4aeaf1f60a9"
},
{
"tags": [
"patch"
],
"url": "https://github.com/vllm-project/vllm/releases/tag/v0.29.1rc0"
},
{
"tags": [
"product"
],
"url": "https://github.com/vllm-project/vllm/"
}
],
"tags": [
"x_open-source"
],
"timeline": [
{
"lang": "en",
"time": "2026-09-30T00:00:00.000Z",
"value": "Advisory disclosed"
},
{
"lang": "en",
"time": "2026-09-30T02:00:00.000Z",
"value": "VulDB entry created"
},
{
"lang": "en",
"time": "2026-09-30T12:42:01.000Z",
"value": "VulDB entry last update"
}
],
"title": "vllm-project vLLM Gemma4UnifiedParser gemma4.rs denial of service",
"x_generator": [
"VulDB PVTS v202609"
]
}
},
"cveMetadata": {
"assignerOrgId": "1af790b2-7ee1-4545-860a-a788eba489b5",
"assignerShortName": "VulDB",
"cveId": "CVE-2026-103241",
"datePublished": "2026-09-30T16:45:13.581Z",
"dateReserved": "2026-09-30T10:36:13.164Z",
"dateUpdated": "2026-09-30T16:45:13.581Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-100654 (GCVE-0-2026-100654)
Vulnerability from cvelistv5 – Published: 2026-09-26 13:23 – Updated: 2026-09-30 14:51
VLAI
EPSS
VEX
Title
vLLM before 0.29.0 Denial of Service via out-of-range stop_token_ids
Summary
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).
Severity
SSVC
Exploitation: none
Automatable: no
Technical Impact: partial
CISA Coordinator · CISA-ADP (v2.0.3)
Decision recorded 2026-09-30 14:51 UTC
CWE
- CWE-129 - Improper Validation of Array Index
Assigner
References
2 references
| URL | Tags |
|---|---|
| https://github.com/vllm-project/vllm/security/adv… | vendor-advisory |
| https://www.vulncheck.com/advisories/vllm-before-… | third-party-advisory |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| vllm-project | vllm |
Affected:
0 , < 0.29.0
(semver)
Unaffected: 0.29.0 (semver) cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:* |
Date Public
2026-09-12 00:00
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-100654",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "no"
},
{
"Technical Impact": "partial"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-09-30T14:51:29.949766Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-09-30T14:51:39.132Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"cna": {
"affected": [
{
"defaultStatus": "unaffected",
"packageURL": "pkg:pypi/vllm",
"product": "vllm",
"vendor": "vllm-project",
"versions": [
{
"lessThan": "0.29.0",
"status": "affected",
"version": "0",
"versionType": "semver"
},
{
"status": "unaffected",
"version": "0.29.0",
"versionType": "semver"
}
]
}
],
"cpeApplicability": [
{
"nodes": [
{
"cpeMatch": [
{
"criteria": "cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:*",
"versionEndExcluding": "0.29.0",
"vulnerable": true
}
],
"negate": false,
"operator": "OR"
}
]
}
],
"credits": [
{
"lang": "en",
"type": "reporter",
"value": "passer12"
},
{
"lang": "en",
"type": "coordinator",
"value": "jperezdealgaba"
},
{
"lang": "en",
"type": "finder",
"value": "QwertyJack"
}
],
"datePublic": "2026-09-12T00:00:00.000Z",
"descriptions": [
{
"lang": "en",
"value": "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 \u003e 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)."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "LOW",
"attackRequirements": "NONE",
"attackVector": "NETWORK",
"baseScore": 7.1,
"baseSeverity": "HIGH",
"privilegesRequired": "LOW",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "NONE",
"vectorString": "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",
"version": "4.0",
"vulnAvailabilityImpact": "HIGH",
"vulnConfidentialityImpact": "NONE",
"vulnIntegrityImpact": "NONE"
},
"format": "CVSS"
},
{
"cvssV3_1": {
"attackComplexity": "LOW",
"attackVector": "NETWORK",
"availabilityImpact": "HIGH",
"baseScore": 6.5,
"baseSeverity": "MEDIUM",
"confidentialityImpact": "NONE",
"integrityImpact": "NONE",
"privilegesRequired": "LOW",
"scope": "UNCHANGED",
"userInteraction": "NONE",
"vectorString": "CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H",
"version": "3.1"
},
"format": "CVSS"
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-129",
"description": "Improper Validation of Array Index",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-09-26T13:23:23.312Z",
"orgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"shortName": "VulnCheck"
},
"references": [
{
"name": "GitHub Security Advisory (GHSA-v5gm-qgmv-gc6c)",
"tags": [
"vendor-advisory"
],
"url": "https://github.com/vllm-project/vllm/security/advisories/GHSA-v5gm-qgmv-gc6c"
},
{
"name": "VulnCheck Advisory: vLLM before 0.29.0 Denial of Service via out-of-range stop_token_ids",
"tags": [
"third-party-advisory"
],
"url": "https://www.vulncheck.com/advisories/vllm-before-0.29.0-denial-of-service-via-out-of-range-stop-token-ids"
}
],
"title": "vLLM before 0.29.0 Denial of Service via out-of-range stop_token_ids",
"x_generator": {
"engine": "vulncheck-endgame"
}
}
},
"cveMetadata": {
"assignerOrgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"assignerShortName": "VulnCheck",
"cveId": "CVE-2026-100654",
"datePublished": "2026-09-26T13:23:23.312Z",
"dateReserved": "2026-09-26T02:33:07.899Z",
"dateUpdated": "2026-09-30T14:51:39.132Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-100653 (GCVE-0-2026-100653)
Vulnerability from cvelistv5 – Published: 2026-09-26 13:23 – Updated: 2026-10-01 11:06
VLAI
EPSS
VEX
Title
vLLM 0.22.1 before 0.28.0 Incomplete Artifact Pin Propagation
Summary
vLLM is an inference and serving engine for large language models. In versions from 0.22.1 through 0.28.0, the operator-supplied model revision pin (--revision / --code-revision) is not propagated to several Hugging Face artifact loads for the FunAudioChat and Tarsier2 architectures: the WhisperFeatureExtractor and speech_tokenizer PreTrainedTokenizerFast loads in vllm/model_executor/models/funaudiochat.py and the Qwen2VLConfig.from_pretrained call used by Tarsier2ProcessingInfo in vllm/model_executor/models/qwen2_vl.py. As a result, deployments pinned to a reviewed revision still resolve these behavior-affecting processor, tokenizer, and config artifacts from the repository's default revision, so a later change to the upstream default branch can alter audio preprocessing, speech tokenizer behavior, or Tarsier2 configuration without any change to the operator's configured pin. This is a supply-chain integrity and reproducibility failure for pinned deployments; it is residual to the earlier fix tracked as GHSA-3ww4-5jv9-j5gm / CVE-2026-47155 and does not constitute remote code execution or a trust_remote_code=False bypass. The issue is fixed in version 0.28.0.
Severity
SSVC
Exploitation: poc
Automatable: no
Technical Impact: partial
CISA Coordinator · CISA-ADP (v2.0.3)
Decision recorded 2026-09-30 17:46 UTC
CWE
- CWE-348 - Use of Less Trusted Source
Assigner
References
3 references
| URL | Tags |
|---|---|
| https://github.com/vllm-project/vllm/security/adv… | vendor-advisory |
| https://github.com/vllm-project/vllm/commit/d26a2… | patch |
| https://www.vulncheck.com/advisories/vllm-0.22.1-… | third-party-advisory |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| vllm-project | vllm |
Affected:
0.22.1 , < 0.28.0
(semver)
Unaffected: 0.28.0 (semver) cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:* |
Date Public
2026-09-12 00:00
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-100653",
"options": [
{
"Exploitation": "poc"
},
{
"Automatable": "no"
},
{
"Technical Impact": "partial"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-09-30T17:46:29.258369Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-09-30T17:46:36.993Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"references": [
{
"tags": [
"exploit"
],
"url": "https://github.com/vllm-project/vllm/security/advisories/GHSA-hhv2-872h-628q"
}
],
"title": "CISA ADP Vulnrichment"
}
],
"cna": {
"affected": [
{
"defaultStatus": "unaffected",
"packageURL": "pkg:pypi/vllm",
"product": "vllm",
"vendor": "vllm-project",
"versions": [
{
"lessThan": "0.28.0",
"status": "affected",
"version": "0.22.1",
"versionType": "semver"
},
{
"status": "unaffected",
"version": "0.28.0",
"versionType": "semver"
}
]
}
],
"cpeApplicability": [
{
"nodes": [
{
"cpeMatch": [
{
"criteria": "cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:*",
"versionEndExcluding": "0.28.0",
"versionStartIncluding": "0.22.1",
"vulnerable": true
}
],
"negate": false,
"operator": "OR"
}
]
}
],
"credits": [
{
"lang": "en",
"type": "reporter",
"value": "rexpository"
},
{
"lang": "en",
"type": "coordinator",
"value": "jperezdealgaba"
},
{
"lang": "en",
"type": "remediation developer",
"value": "KernelClint"
}
],
"datePublic": "2026-09-12T00:00:00.000Z",
"descriptions": [
{
"lang": "en",
"value": "vLLM is an inference and serving engine for large language models. In versions from 0.22.1 through 0.28.0, the operator-supplied model revision pin (--revision / --code-revision) is not propagated to several Hugging Face artifact loads for the FunAudioChat and Tarsier2 architectures: the WhisperFeatureExtractor and speech_tokenizer PreTrainedTokenizerFast loads in vllm/model_executor/models/funaudiochat.py and the Qwen2VLConfig.from_pretrained call used by Tarsier2ProcessingInfo in vllm/model_executor/models/qwen2_vl.py. As a result, deployments pinned to a reviewed revision still resolve these behavior-affecting processor, tokenizer, and config artifacts from the repository\u0027s default revision, so a later change to the upstream default branch can alter audio preprocessing, speech tokenizer behavior, or Tarsier2 configuration without any change to the operator\u0027s configured pin. This is a supply-chain integrity and reproducibility failure for pinned deployments; it is residual to the earlier fix tracked as GHSA-3ww4-5jv9-j5gm / CVE-2026-47155 and does not constitute remote code execution or a trust_remote_code=False bypass. The issue is fixed in version 0.28.0."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "HIGH",
"attackRequirements": "PRESENT",
"attackVector": "NETWORK",
"baseScore": 8.3,
"baseSeverity": "HIGH",
"privilegesRequired": "NONE",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "NONE",
"vectorString": "CVSS:4.0/AV:N/AC:H/AT:P/PR:N/UI:N/VC:L/VI:H/VA:N/SC:N/SI:N/SA:N",
"version": "4.0",
"vulnAvailabilityImpact": "NONE",
"vulnConfidentialityImpact": "LOW",
"vulnIntegrityImpact": "HIGH"
},
"format": "CVSS"
},
{
"cvssV3_1": {
"attackComplexity": "HIGH",
"attackVector": "NETWORK",
"availabilityImpact": "NONE",
"baseScore": 6.5,
"baseSeverity": "MEDIUM",
"confidentialityImpact": "LOW",
"integrityImpact": "HIGH",
"privilegesRequired": "NONE",
"scope": "UNCHANGED",
"userInteraction": "NONE",
"vectorString": "CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:L/I:H/A:N",
"version": "3.1"
},
"format": "CVSS"
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-348",
"description": "Use of Less Trusted Source",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-10-01T11:06:45.139Z",
"orgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"shortName": "VulnCheck"
},
"references": [
{
"name": "GitHub Security Advisory (GHSA-hhv2-872h-628q)",
"tags": [
"vendor-advisory"
],
"url": "https://github.com/vllm-project/vllm/security/advisories/GHSA-hhv2-872h-628q"
},
{
"name": "Patch Commit",
"tags": [
"patch"
],
"url": "https://github.com/vllm-project/vllm/commit/d26a28ab033697f55a1414b5b0435de7cd6045b6"
},
{
"name": "VulnCheck Advisory: vLLM 0.22.1 before 0.28.0 Incomplete Artifact Pin Propagation",
"tags": [
"third-party-advisory"
],
"url": "https://www.vulncheck.com/advisories/vllm-0.22.1-before-0.28.0-incomplete-artifact-pin-propagation"
}
],
"title": "vLLM 0.22.1 before 0.28.0 Incomplete Artifact Pin Propagation",
"x_generator": {
"engine": "vulncheck-endgame"
}
}
},
"cveMetadata": {
"assignerOrgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"assignerShortName": "VulnCheck",
"cveId": "CVE-2026-100653",
"datePublished": "2026-09-26T13:23:22.605Z",
"dateReserved": "2026-09-26T02:33:07.899Z",
"dateUpdated": "2026-10-01T11:06:45.139Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-100652 (GCVE-0-2026-100652)
Vulnerability from cvelistv5 – Published: 2026-09-26 13:23 – Updated: 2026-09-28 16:42
VLAI
EPSS
VEX
Title
vLLM 0.22.0 through 0.23.0 Denial of Service via stop_token_ids
Summary
vLLM versions 0.22.0 through 0.23.0 fail to validate stop_token_ids against vocabulary bounds in Rust HTTP and gRPC frontends, allowing out-of-vocabulary token IDs to reach MinTokensLogitsProcessor. Attackers can submit requests with min_tokens greater than zero and out-of-vocabulary stop_token_ids to trigger CUDA tensor indexing failures that leave EngineCore in a fatal state requiring service restart.
Severity
SSVC
Exploitation: none
Automatable: no
Technical Impact: partial
CISA Coordinator · CISA-ADP (v2.0.3)
Decision recorded 2026-09-28 16:41 UTC
CWE
- CWE-20 - Improper Input Validation
Assigner
References
2 references
| URL | Tags |
|---|---|
| https://github.com/vllm-project/vllm/security/adv… | vendor-advisory |
| https://www.vulncheck.com/advisories/vllm-0.22.0-… | third-party-advisory |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| vllm-project | vllm |
Affected:
0.22.0 , < 0.24.0
(semver)
Unaffected: 0.24.0 (semver) cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:* |
Date Public
2026-09-12 00:00
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-100652",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "no"
},
{
"Technical Impact": "partial"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-09-28T16:41:47.920735Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-09-28T16:42:13.668Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"cna": {
"affected": [
{
"defaultStatus": "unaffected",
"packageURL": "pkg:pypi/vllm",
"product": "vllm",
"vendor": "vllm-project",
"versions": [
{
"lessThan": "0.24.0",
"status": "affected",
"version": "0.22.0",
"versionType": "semver"
},
{
"status": "unaffected",
"version": "0.24.0",
"versionType": "semver"
}
]
}
],
"cpeApplicability": [
{
"nodes": [
{
"cpeMatch": [
{
"criteria": "cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:*",
"versionEndExcluding": "0.24.0",
"versionStartIncluding": "0.22.0",
"vulnerable": true
}
],
"negate": false,
"operator": "OR"
}
]
}
],
"credits": [
{
"lang": "en",
"type": "reporter",
"value": "rexpository"
},
{
"lang": "en",
"type": "coordinator",
"value": "jperezdealgaba"
}
],
"datePublic": "2026-09-12T00:00:00.000Z",
"descriptions": [
{
"lang": "en",
"value": "vLLM versions 0.22.0 through 0.23.0 fail to validate stop_token_ids against vocabulary bounds in Rust HTTP and gRPC frontends, allowing out-of-vocabulary token IDs to reach MinTokensLogitsProcessor. Attackers can submit requests with min_tokens greater than zero and out-of-vocabulary stop_token_ids to trigger CUDA tensor indexing failures that leave EngineCore in a fatal state requiring service restart."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "HIGH",
"attackRequirements": "PRESENT",
"attackVector": "NETWORK",
"baseScore": 8.2,
"baseSeverity": "HIGH",
"privilegesRequired": "NONE",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "NONE",
"vectorString": "CVSS:4.0/AV:N/AC:H/AT:P/PR:N/UI:N/VC:N/VI:N/VA:H/SC:N/SI:N/SA:N",
"version": "4.0",
"vulnAvailabilityImpact": "HIGH",
"vulnConfidentialityImpact": "NONE",
"vulnIntegrityImpact": "NONE"
},
"format": "CVSS"
},
{
"cvssV3_1": {
"attackComplexity": "HIGH",
"attackVector": "NETWORK",
"availabilityImpact": "HIGH",
"baseScore": 5.9,
"baseSeverity": "MEDIUM",
"confidentialityImpact": "NONE",
"integrityImpact": "NONE",
"privilegesRequired": "NONE",
"scope": "UNCHANGED",
"userInteraction": "NONE",
"vectorString": "CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:N/A:H",
"version": "3.1"
},
"format": "CVSS"
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-20",
"description": "Improper Input Validation",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-09-26T13:23:21.865Z",
"orgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"shortName": "VulnCheck"
},
"references": [
{
"name": "GitHub Security Advisory (GHSA-qff2-492f-9fm4)",
"tags": [
"vendor-advisory"
],
"url": "https://github.com/vllm-project/vllm/security/advisories/GHSA-qff2-492f-9fm4"
},
{
"name": "VulnCheck Advisory: vLLM 0.22.0 through 0.23.0 Denial of Service via stop_token_ids",
"tags": [
"third-party-advisory"
],
"url": "https://www.vulncheck.com/advisories/vllm-0.22.0-through-0.23.0-denial-of-service-via-stop-token-ids"
}
],
"title": "vLLM 0.22.0 through 0.23.0 Denial of Service via stop_token_ids",
"x_generator": {
"engine": "vulncheck-endgame"
}
}
},
"cveMetadata": {
"assignerOrgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"assignerShortName": "VulnCheck",
"cveId": "CVE-2026-100652",
"datePublished": "2026-09-26T13:23:21.865Z",
"dateReserved": "2026-09-26T02:33:07.899Z",
"dateUpdated": "2026-09-28T16:42:13.668Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-100651 (GCVE-0-2026-100651)
Vulnerability from cvelistv5 – Published: 2026-09-26 13:23 – Updated: 2026-09-28 18:25
VLAI
EPSS
VEX
Title
vllm before 0.29.0 Denial of Service via Decoder Prompt Length Bypass
Summary
vLLM before 0.29.0 fails to enforce decoder prompt-length validation on the disaggregated serving endpoint /inference/v1/generate. When the request contains a 'features' (multimodal) payload, vllm/entrypoints/serve/disagg/serving.py builds a multimodal EngineInput directly from the caller-supplied token_ids, and GenerateRequest.token_ids (vllm/entrypoints/serve/disagg/protocol.py) is not checked against model_config.max_model_len. For multimodal processors that report skip_prompt_length_check=True (for example Nemotron Parse, Whisper, and FireRedLID), InputProcessor._validate_prompt_len() returns immediately for both encoder and decoder prompts, so an overlong prompt becomes an EngineCoreRequest and reaches the worker input-batch copy into a fixed max_model_len-wide NumPy row. A client able to reach the endpoint on an affected model configuration can therefore submit an overlong token_ids list to trigger a worker failure and denial of service. Fixed in 0.29.0.
Severity
SSVC
Exploitation: poc
Automatable: no
Technical Impact: partial
CISA Coordinator · CISA-ADP (v2.0.3)
Decision recorded 2026-09-28 18:21 UTC
CWE
- CWE-400 - Uncontrolled Resource Consumption
Assigner
References
2 references
| URL | Tags |
|---|---|
| https://github.com/vllm-project/vllm/security/adv… | vendor-advisory |
| https://www.vulncheck.com/advisories/vllm-before-… | third-party-advisory |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| vllm-project | vllm |
Affected:
0 , < 0.29.0
(semver)
Unaffected: 0.29.0 (semver) cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:* |
Date Public
2026-09-12 00:00
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-100651",
"options": [
{
"Exploitation": "poc"
},
{
"Automatable": "no"
},
{
"Technical Impact": "partial"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-09-28T18:21:26.899493Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-09-28T18:25:40.222Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"references": [
{
"tags": [
"exploit"
],
"url": "https://github.com/vllm-project/vllm/security/advisories/GHSA-3mqx-f33v-vgp9"
}
],
"title": "CISA ADP Vulnrichment"
}
],
"cna": {
"affected": [
{
"defaultStatus": "unaffected",
"packageURL": "pkg:pypi/vllm",
"product": "vllm",
"vendor": "vllm-project",
"versions": [
{
"lessThan": "0.29.0",
"status": "affected",
"version": "0",
"versionType": "semver"
},
{
"status": "unaffected",
"version": "0.29.0",
"versionType": "semver"
}
]
}
],
"cpeApplicability": [
{
"nodes": [
{
"cpeMatch": [
{
"criteria": "cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:*",
"versionEndExcluding": "0.29.0",
"vulnerable": true
}
],
"negate": false,
"operator": "OR"
}
]
}
],
"credits": [
{
"lang": "en",
"type": "reporter",
"value": "rexpository"
},
{
"lang": "en",
"type": "coordinator",
"value": "jperezdealgaba"
}
],
"datePublic": "2026-09-12T00:00:00.000Z",
"descriptions": [
{
"lang": "en",
"value": "vLLM before 0.29.0 fails to enforce decoder prompt-length validation on the disaggregated serving endpoint /inference/v1/generate. When the request contains a \u0027features\u0027 (multimodal) payload, vllm/entrypoints/serve/disagg/serving.py builds a multimodal EngineInput directly from the caller-supplied token_ids, and GenerateRequest.token_ids (vllm/entrypoints/serve/disagg/protocol.py) is not checked against model_config.max_model_len. For multimodal processors that report skip_prompt_length_check=True (for example Nemotron Parse, Whisper, and FireRedLID), InputProcessor._validate_prompt_len() returns immediately for both encoder and decoder prompts, so an overlong prompt becomes an EngineCoreRequest and reaches the worker input-batch copy into a fixed max_model_len-wide NumPy row. A client able to reach the endpoint on an affected model configuration can therefore submit an overlong token_ids list to trigger a worker failure and denial of service. Fixed in 0.29.0."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "LOW",
"attackRequirements": "NONE",
"attackVector": "NETWORK",
"baseScore": 7.1,
"baseSeverity": "HIGH",
"privilegesRequired": "LOW",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "NONE",
"vectorString": "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",
"version": "4.0",
"vulnAvailabilityImpact": "HIGH",
"vulnConfidentialityImpact": "NONE",
"vulnIntegrityImpact": "NONE"
},
"format": "CVSS"
},
{
"cvssV3_1": {
"attackComplexity": "LOW",
"attackVector": "NETWORK",
"availabilityImpact": "HIGH",
"baseScore": 6.5,
"baseSeverity": "MEDIUM",
"confidentialityImpact": "NONE",
"integrityImpact": "NONE",
"privilegesRequired": "LOW",
"scope": "UNCHANGED",
"userInteraction": "NONE",
"vectorString": "CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H",
"version": "3.1"
},
"format": "CVSS"
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-400",
"description": "Uncontrolled Resource Consumption",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-09-26T13:23:21.141Z",
"orgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"shortName": "VulnCheck"
},
"references": [
{
"name": "GitHub Security Advisory (GHSA-3mqx-f33v-vgp9)",
"tags": [
"vendor-advisory"
],
"url": "https://github.com/vllm-project/vllm/security/advisories/GHSA-3mqx-f33v-vgp9"
},
{
"name": "VulnCheck Advisory: vllm before 0.29.0 Denial of Service via Decoder Prompt Length Bypass",
"tags": [
"third-party-advisory"
],
"url": "https://www.vulncheck.com/advisories/vllm-before-0.29.0-denial-of-service-via-decoder-prompt-length-bypass"
}
],
"title": "vllm before 0.29.0 Denial of Service via Decoder Prompt Length Bypass",
"x_generator": {
"engine": "vulncheck-endgame"
}
}
},
"cveMetadata": {
"assignerOrgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"assignerShortName": "VulnCheck",
"cveId": "CVE-2026-100651",
"datePublished": "2026-09-26T13:23:21.141Z",
"dateReserved": "2026-09-26T02:33:07.899Z",
"dateUpdated": "2026-09-28T18:25:40.222Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-100650 (GCVE-0-2026-100650)
Vulnerability from cvelistv5 – Published: 2026-09-26 13:23 – Updated: 2026-10-01 11:06
VLAI
EPSS
VEX
Title
vLLM before 0.29.0 Resource Exhaustion via Unbounded Media Materialization
Summary
vLLM through 0.29.0 fetches and fully materializes remote or inline media before enforcing its documented media controls (the VLLM_MAX_AUDIO_CLIP_FILESIZE_MB compressed-audio size cap, default 25 MB, and the per-modality --limit-mm-per-prompt item limits). Across four ingress paths — the shared media-acquisition layer (HTTPConnection.get_bytes()/async_get_bytes()), the chat completions audio_url/base64 path, the batch speech runner, and the Rust frontend POST /tokenize route — the server reads the entire HTTP response body, base64-decodes the inline payload, or spawns one fetch/decode task per media part, and only then applies the limit (or, on some paths, never applies it). A remote attacker can therefore cause the API server or batch-runner process to allocate memory and consume outbound bandwidth proportional to an attacker-chosen body size or media item count before the request is rejected, resulting in pre-inference memory and bandwidth exhaustion (denial of service). The chat and batch surfaces require an API key when one is configured; the Rust frontend /tokenize route is unauthenticated by design. There is no code execution or data disclosure impact.
Severity
SSVC
Exploitation: poc
Automatable: no
Technical Impact: partial
CISA Coordinator · CISA-ADP (v2.0.3)
Decision recorded 2026-09-30 14:49 UTC
CWE
- CWE-400 - Uncontrolled Resource Consumption
Assigner
References
3 references
| URL | Tags |
|---|---|
| https://github.com/vllm-project/vllm/security/adv… | vendor-advisory |
| https://github.com/vllm-project/vllm/commit/752a3… | patch |
| https://www.vulncheck.com/advisories/vllm-before-… | third-party-advisory |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| vllm-project | vllm |
Affected:
0 , < 0.29.0
(semver)
Unaffected: 0.29.0 (semver) cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:* |
Date Public
2026-09-12 00:00
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-100650",
"options": [
{
"Exploitation": "poc"
},
{
"Automatable": "no"
},
{
"Technical Impact": "partial"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-09-30T14:49:40.322864Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-09-30T14:50:14.406Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"references": [
{
"tags": [
"exploit"
],
"url": "https://github.com/vllm-project/vllm/security/advisories/GHSA-p6g9-7v3x-m8mv"
}
],
"title": "CISA ADP Vulnrichment"
}
],
"cna": {
"affected": [
{
"defaultStatus": "unaffected",
"packageURL": "pkg:pypi/vllm",
"product": "vllm",
"vendor": "vllm-project",
"versions": [
{
"lessThan": "0.29.0",
"status": "affected",
"version": "0",
"versionType": "semver"
},
{
"status": "unaffected",
"version": "0.29.0",
"versionType": "semver"
}
]
}
],
"cpeApplicability": [
{
"nodes": [
{
"cpeMatch": [
{
"criteria": "cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:*",
"versionEndExcluding": "0.29.0",
"vulnerable": true
}
],
"negate": false,
"operator": "OR"
}
]
}
],
"credits": [
{
"lang": "en",
"type": "finder",
"value": "KernelClint"
},
{
"lang": "en",
"type": "finder",
"value": "dhalf"
},
{
"lang": "en",
"type": "coordinator",
"value": "jperezdealgaba"
}
],
"datePublic": "2026-09-12T00:00:00.000Z",
"descriptions": [
{
"lang": "en",
"value": "vLLM through 0.29.0 fetches and fully materializes remote or inline media before enforcing its documented media controls (the VLLM_MAX_AUDIO_CLIP_FILESIZE_MB compressed-audio size cap, default 25 MB, and the per-modality --limit-mm-per-prompt item limits). Across four ingress paths \u2014 the shared media-acquisition layer (HTTPConnection.get_bytes()/async_get_bytes()), the chat completions audio_url/base64 path, the batch speech runner, and the Rust frontend POST /tokenize route \u2014 the server reads the entire HTTP response body, base64-decodes the inline payload, or spawns one fetch/decode task per media part, and only then applies the limit (or, on some paths, never applies it). A remote attacker can therefore cause the API server or batch-runner process to allocate memory and consume outbound bandwidth proportional to an attacker-chosen body size or media item count before the request is rejected, resulting in pre-inference memory and bandwidth exhaustion (denial of service). The chat and batch surfaces require an API key when one is configured; the Rust frontend /tokenize route is unauthenticated by design. There is no code execution or data disclosure impact."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "LOW",
"attackRequirements": "NONE",
"attackVector": "NETWORK",
"baseScore": 7.1,
"baseSeverity": "HIGH",
"privilegesRequired": "LOW",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "NONE",
"vectorString": "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",
"version": "4.0",
"vulnAvailabilityImpact": "HIGH",
"vulnConfidentialityImpact": "NONE",
"vulnIntegrityImpact": "NONE"
},
"format": "CVSS"
},
{
"cvssV3_1": {
"attackComplexity": "LOW",
"attackVector": "NETWORK",
"availabilityImpact": "HIGH",
"baseScore": 6.5,
"baseSeverity": "MEDIUM",
"confidentialityImpact": "NONE",
"integrityImpact": "NONE",
"privilegesRequired": "LOW",
"scope": "UNCHANGED",
"userInteraction": "NONE",
"vectorString": "CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H",
"version": "3.1"
},
"format": "CVSS"
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-400",
"description": "Uncontrolled Resource Consumption",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-10-01T11:06:44.741Z",
"orgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"shortName": "VulnCheck"
},
"references": [
{
"name": "GitHub Security Advisory (GHSA-p6g9-7v3x-m8mv)",
"tags": [
"vendor-advisory"
],
"url": "https://github.com/vllm-project/vllm/security/advisories/GHSA-p6g9-7v3x-m8mv"
},
{
"name": "Patch Commit",
"tags": [
"patch"
],
"url": "https://github.com/vllm-project/vllm/commit/752a3a504485790a2e8491cacbb35c137339ad34"
},
{
"name": "VulnCheck Advisory: vLLM before 0.29.0 Resource Exhaustion via Unbounded Media Materialization",
"tags": [
"third-party-advisory"
],
"url": "https://www.vulncheck.com/advisories/vllm-before-0.29.0-resource-exhaustion-via-unbounded-media-materialization"
}
],
"title": "vLLM before 0.29.0 Resource Exhaustion via Unbounded Media Materialization",
"x_generator": {
"engine": "vulncheck-endgame"
}
}
},
"cveMetadata": {
"assignerOrgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"assignerShortName": "VulnCheck",
"cveId": "CVE-2026-100650",
"datePublished": "2026-09-26T13:23:20.418Z",
"dateReserved": "2026-09-26T02:33:07.898Z",
"dateUpdated": "2026-10-01T11:06:44.741Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-100649 (GCVE-0-2026-100649)
Vulnerability from cvelistv5 – Published: 2026-09-26 13:23 – Updated: 2026-09-26 13:23
VLAI
EPSS
VEX
Title
vLLM before 0.29.0 Resource Limit Bypass via Sampler Subclass
Summary
vLLM before 0.29.0 contains a resource-limit bypass vulnerability in PyNvVideoCodec decoder allocation where sampler subclass shadowing allows independent counter increments. Unauthenticated attackers can select different sampler subclasses in video requests to exceed configured decoder limits and exhaust unaccounted GPU memory.
Severity
CWE
- CWE-770 - Allocation of Resources Without Limits or Throttling
Assigner
References
2 references
| URL | Tags |
|---|---|
| https://github.com/vllm-project/vllm/security/adv… | vendor-advisory |
| https://www.vulncheck.com/advisories/vllm-before-… | third-party-advisory |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| vllm-project | vllm |
Affected:
0 , < 0.29.0
(semver)
Unaffected: 0.29.0 (semver) cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:* |
Date Public
2026-09-12 00:00
{
"containers": {
"cna": {
"affected": [
{
"defaultStatus": "unaffected",
"packageURL": "pkg:pypi/vllm",
"product": "vllm",
"vendor": "vllm-project",
"versions": [
{
"lessThan": "0.29.0",
"status": "affected",
"version": "0",
"versionType": "semver"
},
{
"status": "unaffected",
"version": "0.29.0",
"versionType": "semver"
}
]
}
],
"cpeApplicability": [
{
"nodes": [
{
"cpeMatch": [
{
"criteria": "cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:*",
"versionEndExcluding": "0.29.0",
"vulnerable": true
}
],
"negate": false,
"operator": "OR"
}
]
}
],
"credits": [
{
"lang": "en",
"type": "reporter",
"value": "JPengLi"
},
{
"lang": "en",
"type": "coordinator",
"value": "jperezdealgaba"
}
],
"datePublic": "2026-09-12T00:00:00.000Z",
"descriptions": [
{
"lang": "en",
"value": "vLLM before 0.29.0 contains a resource-limit bypass vulnerability in PyNvVideoCodec decoder allocation where sampler subclass shadowing allows independent counter increments. Unauthenticated attackers can select different sampler subclasses in video requests to exceed configured decoder limits and exhaust unaccounted GPU memory."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "HIGH",
"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:H/AT:P/PR:N/UI:N/VC:N/VI:N/VA:L/SC:N/SI:N/SA:N",
"version": "4.0",
"vulnAvailabilityImpact": "LOW",
"vulnConfidentialityImpact": "NONE",
"vulnIntegrityImpact": "NONE"
},
"format": "CVSS"
},
{
"cvssV3_1": {
"attackComplexity": "HIGH",
"attackVector": "NETWORK",
"availabilityImpact": "LOW",
"baseScore": 3.7,
"baseSeverity": "LOW",
"confidentialityImpact": "NONE",
"integrityImpact": "NONE",
"privilegesRequired": "NONE",
"scope": "UNCHANGED",
"userInteraction": "NONE",
"vectorString": "CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:N/A:L",
"version": "3.1"
},
"format": "CVSS"
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-770",
"description": "Allocation of Resources Without Limits or Throttling",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-09-26T13:23:19.687Z",
"orgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"shortName": "VulnCheck"
},
"references": [
{
"name": "GitHub Security Advisory (GHSA-j682-9xp5-rrf3)",
"tags": [
"vendor-advisory"
],
"url": "https://github.com/vllm-project/vllm/security/advisories/GHSA-j682-9xp5-rrf3"
},
{
"name": "VulnCheck Advisory: vLLM before 0.29.0 Resource Limit Bypass via Sampler Subclass",
"tags": [
"third-party-advisory"
],
"url": "https://www.vulncheck.com/advisories/vllm-before-0.29.0-resource-limit-bypass-via-sampler-subclass"
}
],
"title": "vLLM before 0.29.0 Resource Limit Bypass via Sampler Subclass",
"x_generator": {
"engine": "vulncheck-endgame"
}
}
},
"cveMetadata": {
"assignerOrgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"assignerShortName": "VulnCheck",
"cveId": "CVE-2026-100649",
"datePublished": "2026-09-26T13:23:19.687Z",
"dateReserved": "2026-09-26T02:33:07.898Z",
"dateUpdated": "2026-09-26T13:23:19.687Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-100648 (GCVE-0-2026-100648)
Vulnerability from cvelistv5 – Published: 2026-09-26 13:23 – Updated: 2026-09-28 16:40
VLAI
EPSS
VEX
Title
vllm before 0.29.0 Uncontrolled Resource Consumption via Audio Decoding
Summary
vllm before 0.29.0 fails to enforce VLLM_MAX_AUDIO_CLIP_FILESIZE_MB limit in multimodal chat audio decoding, allowing unauthenticated clients to bypass file size restrictions. Attackers can submit oversized audio files through chat endpoints to consume excessive memory and CPU resources during decoding.
Severity
5.3 (Medium)
SSVC
Exploitation: none
Automatable: yes
Technical Impact: partial
CISA Coordinator · CISA-ADP (v2.0.3)
Decision recorded 2026-09-28 16:40 UTC
CWE
- CWE-400 - Uncontrolled Resource Consumption
Assigner
References
2 references
| URL | Tags |
|---|---|
| https://github.com/vllm-project/vllm/security/adv… | vendor-advisory |
| https://www.vulncheck.com/advisories/vllm-before-… | third-party-advisory |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| vllm-project | vllm |
Affected:
0 , < 0.29.0
(semver)
Unaffected: 0.29.0 (semver) cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:* |
Date Public
2026-09-12 00:00
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-100648",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "yes"
},
{
"Technical Impact": "partial"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-09-28T16:40:22.871965Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-09-28T16:40:42.006Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"cna": {
"affected": [
{
"defaultStatus": "unaffected",
"packageURL": "pkg:pypi/vllm",
"product": "vllm",
"vendor": "vllm-project",
"versions": [
{
"lessThan": "0.29.0",
"status": "affected",
"version": "0",
"versionType": "semver"
},
{
"status": "unaffected",
"version": "0.29.0",
"versionType": "semver"
}
]
}
],
"cpeApplicability": [
{
"nodes": [
{
"cpeMatch": [
{
"criteria": "cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:*",
"versionEndExcluding": "0.29.0",
"vulnerable": true
}
],
"negate": false,
"operator": "OR"
}
]
}
],
"credits": [
{
"lang": "en",
"type": "reporter",
"value": "santhreal"
},
{
"lang": "en",
"type": "coordinator",
"value": "jperezdealgaba"
}
],
"datePublic": "2026-09-12T00:00:00.000Z",
"descriptions": [
{
"lang": "en",
"value": "vllm before 0.29.0 fails to enforce VLLM_MAX_AUDIO_CLIP_FILESIZE_MB limit in multimodal chat audio decoding, allowing unauthenticated clients to bypass file size restrictions. Attackers can submit oversized audio files through chat endpoints to consume excessive memory and CPU resources during decoding."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "LOW",
"attackRequirements": "NONE",
"attackVector": "NETWORK",
"baseScore": 6.9,
"baseSeverity": "MEDIUM",
"privilegesRequired": "NONE",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "NONE",
"vectorString": "CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:N/VI:N/VA:L/SC:N/SI:N/SA:N",
"version": "4.0",
"vulnAvailabilityImpact": "LOW",
"vulnConfidentialityImpact": "NONE",
"vulnIntegrityImpact": "NONE"
},
"format": "CVSS"
},
{
"cvssV3_1": {
"attackComplexity": "LOW",
"attackVector": "NETWORK",
"availabilityImpact": "LOW",
"baseScore": 5.3,
"baseSeverity": "MEDIUM",
"confidentialityImpact": "NONE",
"integrityImpact": "NONE",
"privilegesRequired": "NONE",
"scope": "UNCHANGED",
"userInteraction": "NONE",
"vectorString": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:L",
"version": "3.1"
},
"format": "CVSS"
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-400",
"description": "Uncontrolled Resource Consumption",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-09-26T13:23:18.965Z",
"orgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"shortName": "VulnCheck"
},
"references": [
{
"name": "GitHub Security Advisory (GHSA-jcq2-4gch-5qhf)",
"tags": [
"vendor-advisory"
],
"url": "https://github.com/vllm-project/vllm/security/advisories/GHSA-jcq2-4gch-5qhf"
},
{
"name": "VulnCheck Advisory: vllm before 0.29.0 Uncontrolled Resource Consumption via Audio Decoding",
"tags": [
"third-party-advisory"
],
"url": "https://www.vulncheck.com/advisories/vllm-before-0.29.0-uncontrolled-resource-consumption-via-audio-decoding"
}
],
"title": "vllm before 0.29.0 Uncontrolled Resource Consumption via Audio Decoding",
"x_generator": {
"engine": "vulncheck-endgame"
}
}
},
"cveMetadata": {
"assignerOrgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"assignerShortName": "VulnCheck",
"cveId": "CVE-2026-100648",
"datePublished": "2026-09-26T13:23:18.965Z",
"dateReserved": "2026-09-26T02:33:07.898Z",
"dateUpdated": "2026-09-28T16:40:42.006Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-100647 (GCVE-0-2026-100647)
Vulnerability from cvelistv5 – Published: 2026-09-26 13:23 – Updated: 2026-09-28 18:28
VLAI
EPSS
VEX
Title
vLLM before 0.29.0 CPU Exhaustion via unbounded cache_salt
Summary
vLLM versions before 0.29.0 contain a denial-of-service vulnerability in the cache_salt parameter accepted on OpenAI-compatible and Anthropic API endpoints, which lacks maximum length validation and is processed on the single EngineCore scheduler thread. Unauthenticated attackers can send HTTP requests with multi-hundred-megabyte salt values that trigger expensive pickle serialization and SHA-256 hashing, stalling the scheduler thread and denying service to all concurrent requests.
Severity
5.3 (Medium)
SSVC
Exploitation: none
Automatable: yes
Technical Impact: partial
CISA Coordinator · CISA-ADP (v2.0.3)
Decision recorded 2026-09-28 18:26 UTC
CWE
- CWE-20 - Improper Input Validation
Assigner
References
2 references
| URL | Tags |
|---|---|
| https://github.com/vllm-project/vllm/security/adv… | vendor-advisory |
| https://www.vulncheck.com/advisories/vllm-before-… | third-party-advisory |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| vllm-project | vllm |
Affected:
0 , < 0.29.0
(semver)
Unaffected: 0.29.0 (semver) cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:* |
Date Public
2026-09-12 00:00
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-100647",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "yes"
},
{
"Technical Impact": "partial"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-09-28T18:26:37.598538Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-09-28T18:28:14.772Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"cna": {
"affected": [
{
"defaultStatus": "unaffected",
"packageURL": "pkg:pypi/vllm",
"product": "vllm",
"vendor": "vllm-project",
"versions": [
{
"lessThan": "0.29.0",
"status": "affected",
"version": "0",
"versionType": "semver"
},
{
"status": "unaffected",
"version": "0.29.0",
"versionType": "semver"
}
]
}
],
"cpeApplicability": [
{
"nodes": [
{
"cpeMatch": [
{
"criteria": "cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:*",
"versionEndExcluding": "0.29.0",
"vulnerable": true
}
],
"negate": false,
"operator": "OR"
}
]
}
],
"credits": [
{
"lang": "en",
"type": "reporter",
"value": "wildoranges"
},
{
"lang": "en",
"type": "coordinator",
"value": "jperezdealgaba"
}
],
"datePublic": "2026-09-12T00:00:00.000Z",
"descriptions": [
{
"lang": "en",
"value": "vLLM versions before 0.29.0 contain a denial-of-service vulnerability in the cache_salt parameter accepted on OpenAI-compatible and Anthropic API endpoints, which lacks maximum length validation and is processed on the single EngineCore scheduler thread. Unauthenticated attackers can send HTTP requests with multi-hundred-megabyte salt values that trigger expensive pickle serialization and SHA-256 hashing, stalling the scheduler thread and denying service to all concurrent requests."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "LOW",
"attackRequirements": "NONE",
"attackVector": "NETWORK",
"baseScore": 6.9,
"baseSeverity": "MEDIUM",
"privilegesRequired": "NONE",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "NONE",
"vectorString": "CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:N/VI:N/VA:L/SC:N/SI:N/SA:N",
"version": "4.0",
"vulnAvailabilityImpact": "LOW",
"vulnConfidentialityImpact": "NONE",
"vulnIntegrityImpact": "NONE"
},
"format": "CVSS"
},
{
"cvssV3_1": {
"attackComplexity": "LOW",
"attackVector": "NETWORK",
"availabilityImpact": "LOW",
"baseScore": 5.3,
"baseSeverity": "MEDIUM",
"confidentialityImpact": "NONE",
"integrityImpact": "NONE",
"privilegesRequired": "NONE",
"scope": "UNCHANGED",
"userInteraction": "NONE",
"vectorString": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:L",
"version": "3.1"
},
"format": "CVSS"
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-20",
"description": "Improper Input Validation",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-09-26T13:23:18.240Z",
"orgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"shortName": "VulnCheck"
},
"references": [
{
"name": "GitHub Security Advisory (GHSA-wpww-v874-ph2p)",
"tags": [
"vendor-advisory"
],
"url": "https://github.com/vllm-project/vllm/security/advisories/GHSA-wpww-v874-ph2p"
},
{
"name": "VulnCheck Advisory: vLLM before 0.29.0 CPU Exhaustion via unbounded cache_salt",
"tags": [
"third-party-advisory"
],
"url": "https://www.vulncheck.com/advisories/vllm-before-0.29.0-cpu-exhaustion-via-unbounded-cache-salt"
}
],
"title": "vLLM before 0.29.0 CPU Exhaustion via unbounded cache_salt",
"x_generator": {
"engine": "vulncheck-endgame"
}
}
},
"cveMetadata": {
"assignerOrgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"assignerShortName": "VulnCheck",
"cveId": "CVE-2026-100647",
"datePublished": "2026-09-26T13:23:18.240Z",
"dateReserved": "2026-09-26T02:33:07.898Z",
"dateUpdated": "2026-09-28T18:28:14.772Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-94627 (GCVE-0-2026-94627)
Vulnerability from cvelistv5 – Published: 2026-09-21 22:04 – Updated: 2026-09-22 12:51
VLAI
EPSS
VEX
Title
vLLM through 0.29.0 GPU KV Cache Leak via Mooncake Transfer ID Collision
Summary
vLLM Mooncake connector through 0.29.0 fails to properly manage GPU KV cache block ownership when concurrent child requests share a single transfer ID in prefill/decode disaggregated deployments. Attackers can trigger GPU memory exhaustion by submitting completion requests with multiple prompts, causing orphaned KV cache blocks to accumulate until process restart and eventually preventing legitimate requests from executing.
Severity
SSVC
Exploitation: none
Automatable: yes
Technical Impact: partial
CISA Coordinator · CISA-ADP (v2.0.3)
Decision recorded 2026-09-22 12:50 UTC
CWE
- CWE-401 - Missing Release of Memory after Effective Lifetime
Assigner
References
4 references
| URL | Tags |
|---|---|
| https://github.com/vllm-project/vllm/pull/49796 | issue-trackingpatch |
| https://github.com/vllm-project/vllm/blob/v0.29.0… | technical-description |
| https://github.com/vllm-project/vllm | product |
| https://www.vulncheck.com/advisories/vllm-through… | third-party-advisory |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| vllm-project | vllm |
Affected:
0 , ≤ 0.29.0
(semver)
cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:* |
Date Public
2026-07-25 00:00
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-94627",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "yes"
},
{
"Technical Impact": "partial"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-09-22T12:50:58.697049Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-09-22T12:51:23.062Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"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": "finder",
"value": "Mingkai Yu"
},
{
"lang": "en",
"type": "finder",
"value": "Jiapeng Li"
},
{
"lang": "en",
"type": "finder",
"value": "Jiajia Liu"
}
],
"datePublic": "2026-07-25T00:00:00.000Z",
"descriptions": [
{
"lang": "en",
"value": "vLLM Mooncake connector through 0.29.0 fails to properly manage GPU KV cache block ownership when concurrent child requests share a single transfer ID in prefill/decode disaggregated deployments. Attackers can trigger GPU memory exhaustion by submitting completion requests with multiple prompts, causing orphaned KV cache blocks to accumulate until process restart and eventually preventing legitimate requests from executing."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "LOW",
"attackRequirements": "NONE",
"attackVector": "NETWORK",
"baseScore": 8.7,
"baseSeverity": "HIGH",
"privilegesRequired": "NONE",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "NONE",
"vectorString": "CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:N/VI:N/VA:H/SC:N/SI:N/SA:N",
"version": "4.0",
"vulnAvailabilityImpact": "HIGH",
"vulnConfidentialityImpact": "NONE",
"vulnIntegrityImpact": "NONE"
},
"format": "CVSS"
},
{
"cvssV3_1": {
"attackComplexity": "LOW",
"attackVector": "NETWORK",
"availabilityImpact": "HIGH",
"baseScore": 7.5,
"baseSeverity": "HIGH",
"confidentialityImpact": "NONE",
"integrityImpact": "NONE",
"privilegesRequired": "NONE",
"scope": "UNCHANGED",
"userInteraction": "NONE",
"vectorString": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H",
"version": "3.1"
},
"format": "CVSS"
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-401",
"description": "Missing Release of Memory after Effective Lifetime",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-09-21T22:04:16.559Z",
"orgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"shortName": "VulnCheck"
},
"references": [
{
"name": "Pull Request #49796",
"tags": [
"issue-tracking",
"patch"
],
"url": "https://github.com/vllm-project/vllm/pull/49796"
},
{
"tags": [
"technical-description"
],
"url": "https://github.com/vllm-project/vllm/blob/v0.29.0/vllm/distributed/kv_transfer/kv_connector/v1/mooncake/mooncake_connector.py#L1978-L1989"
},
{
"tags": [
"product"
],
"url": "https://github.com/vllm-project/vllm"
},
{
"name": "VulnCheck Advisory: vLLM through 0.29.0 GPU KV Cache Leak via Mooncake Transfer ID Collision",
"tags": [
"third-party-advisory"
],
"url": "https://www.vulncheck.com/advisories/vllm-through-0.29.0-gpu-kv-cache-leak-via-mooncake-transfer-id-collision"
}
],
"title": "vLLM through 0.29.0 GPU KV Cache Leak via Mooncake Transfer ID Collision",
"x_generator": {
"engine": "vulncheck-endgame"
}
}
},
"cveMetadata": {
"assignerOrgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"assignerShortName": "VulnCheck",
"cveId": "CVE-2026-94627",
"datePublished": "2026-09-21T22:04:16.559Z",
"dateReserved": "2026-09-21T21:42:30.080Z",
"dateUpdated": "2026-09-22T12:51:23.062Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-94626 (GCVE-0-2026-94626)
Vulnerability from cvelistv5 – Published: 2026-09-21 22:04 – Updated: 2026-09-24 22:54
VLAI
EPSS
VEX
Title
vLLM through 0.29.0 Memory Exhaustion via Unvalidated NIXL tp_size
Summary
vLLM through 0.29.0 fails to validate the tp_size parameter in kv_transfer_params on OpenAI-compatible completion endpoints, allowing attackers to allocate unbounded memory. Attackers can supply arbitrary tp_size values in prefill/decode disaggregated deployments to exhaust memory and trigger kernel OOM-kill of the decode worker process.
Severity
SSVC
Exploitation: none
Automatable: yes
Technical Impact: partial
CISA Coordinator · CISA-ADP (v2.0.3)
Decision recorded 2026-09-24 22:54 UTC
CWE
- CWE-789 - Memory Allocation with Excessive Size Value
Assigner
References
5 references
| URL | Tags |
|---|---|
| https://github.com/vllm-project/vllm/pull/51137 | issue-trackingpatch |
| 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://github.com/vllm-project/vllm | product |
| https://www.vulncheck.com/advisories/vllm-through… | third-party-advisory |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| vllm-project | vllm |
Affected:
0 , ≤ 0.29.0
(semver)
cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:* |
Date Public
2026-08-05 00:00
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-94626",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "yes"
},
{
"Technical Impact": "partial"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-09-24T22:54:32.814414Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-09-24T22:54:41.241Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"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": "finder",
"value": "Mingkai Yu"
},
{
"lang": "en",
"type": "finder",
"value": "Jiapeng Li"
},
{
"lang": "en",
"type": "finder",
"value": "Jiajia Liu"
}
],
"datePublic": "2026-08-05T00:00:00.000Z",
"descriptions": [
{
"lang": "en",
"value": "vLLM through 0.29.0 fails to validate the tp_size parameter in kv_transfer_params on OpenAI-compatible completion endpoints, allowing attackers to allocate unbounded memory. Attackers can supply arbitrary tp_size values in prefill/decode disaggregated deployments to exhaust memory and trigger kernel OOM-kill of the decode worker process."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "LOW",
"attackRequirements": "NONE",
"attackVector": "NETWORK",
"baseScore": 8.7,
"baseSeverity": "HIGH",
"privilegesRequired": "NONE",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "NONE",
"vectorString": "CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:N/VI:N/VA:H/SC:N/SI:N/SA:N",
"version": "4.0",
"vulnAvailabilityImpact": "HIGH",
"vulnConfidentialityImpact": "NONE",
"vulnIntegrityImpact": "NONE"
},
"format": "CVSS"
},
{
"cvssV3_1": {
"attackComplexity": "LOW",
"attackVector": "NETWORK",
"availabilityImpact": "HIGH",
"baseScore": 7.5,
"baseSeverity": "HIGH",
"confidentialityImpact": "NONE",
"integrityImpact": "NONE",
"privilegesRequired": "NONE",
"scope": "UNCHANGED",
"userInteraction": "NONE",
"vectorString": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H",
"version": "3.1"
},
"format": "CVSS"
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-789",
"description": "Memory Allocation with Excessive Size Value",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-09-21T22:04:15.352Z",
"orgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"shortName": "VulnCheck"
},
"references": [
{
"name": "Pull Request #51137",
"tags": [
"issue-tracking",
"patch"
],
"url": "https://github.com/vllm-project/vllm/pull/51137"
},
{
"tags": [
"technical-description"
],
"url": "https://github.com/vllm-project/vllm/blob/v0.29.0/vllm/distributed/kv_transfer/kv_connector/utils.py#L569-L573"
},
{
"tags": [
"technical-description"
],
"url": "https://github.com/vllm-project/vllm/blob/v0.29.0/vllm/distributed/kv_transfer/kv_connector/v1/nixl/metadata.py#L277"
},
{
"tags": [
"product"
],
"url": "https://github.com/vllm-project/vllm"
},
{
"name": "VulnCheck Advisory: vLLM through 0.29.0 Memory Exhaustion via Unvalidated NIXL tp_size",
"tags": [
"third-party-advisory"
],
"url": "https://www.vulncheck.com/advisories/vllm-through-0.29.0-memory-exhaustion-via-unvalidated-nixl-tp-size"
}
],
"title": "vLLM through 0.29.0 Memory Exhaustion via Unvalidated NIXL tp_size",
"x_generator": {
"engine": "vulncheck-endgame"
}
}
},
"cveMetadata": {
"assignerOrgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"assignerShortName": "VulnCheck",
"cveId": "CVE-2026-94626",
"datePublished": "2026-09-21T22:04:15.352Z",
"dateReserved": "2026-09-21T21:42:28.781Z",
"dateUpdated": "2026-09-24T22:54:41.241Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-94625 (GCVE-0-2026-94625)
Vulnerability from cvelistv5 – Published: 2026-09-21 22:04 – Updated: 2026-09-23 15:09
VLAI
EPSS
VEX
Title
vLLM through 0.29.0 Resource Exhaustion via Ownerless Mooncake Transfer Placeholders
Summary
vLLM through 0.29.0 contains a resource exhaustion vulnerability in MooncakeConnector where rejected prefill requests create ownerless transfer placeholders that are never reclaimed. Attackers can send rejected requests to exhaust sender task pools, causing valid requests to be delayed by up to 480 seconds while health checks continue returning success.
Severity
5.3 (Medium)
SSVC
Exploitation: none
Automatable: yes
Technical Impact: partial
CISA Coordinator · CISA-ADP (v2.0.3)
Decision recorded 2026-09-23 15:09 UTC
CWE
- CWE-772 - Missing Release of Resource after Effective Lifetime
Assigner
References
4 references
| URL | Tags |
|---|---|
| https://github.com/vllm-project/vllm/pull/51236 | issue-trackingpatch |
| https://github.com/vllm-project/vllm/blob/v0.29.0… | technical-description |
| https://github.com/vllm-project/vllm | product |
| https://www.vulncheck.com/advisories/vllm-through… | third-party-advisory |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| vllm-project | vllm |
Affected:
0 , ≤ 0.29.0
(semver)
cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:* |
Date Public
2026-08-06 00:00
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-94625",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "yes"
},
{
"Technical Impact": "partial"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-09-23T15:09:07.861181Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-09-23T15:09:16.926Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"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": "finder",
"value": "Mingkai Yu"
},
{
"lang": "en",
"type": "finder",
"value": "Jiapeng Li"
},
{
"lang": "en",
"type": "finder",
"value": "Jiajia Liu"
}
],
"datePublic": "2026-08-06T00:00:00.000Z",
"descriptions": [
{
"lang": "en",
"value": "vLLM through 0.29.0 contains a resource exhaustion vulnerability in MooncakeConnector where rejected prefill requests create ownerless transfer placeholders that are never reclaimed. Attackers can send rejected requests to exhaust sender task pools, causing valid requests to be delayed by up to 480 seconds while health checks continue returning success."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "LOW",
"attackRequirements": "NONE",
"attackVector": "NETWORK",
"baseScore": 6.9,
"baseSeverity": "MEDIUM",
"privilegesRequired": "NONE",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "NONE",
"vectorString": "CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:N/VI:N/VA:L/SC:N/SI:N/SA:N",
"version": "4.0",
"vulnAvailabilityImpact": "LOW",
"vulnConfidentialityImpact": "NONE",
"vulnIntegrityImpact": "NONE"
},
"format": "CVSS"
},
{
"cvssV3_1": {
"attackComplexity": "LOW",
"attackVector": "NETWORK",
"availabilityImpact": "LOW",
"baseScore": 5.3,
"baseSeverity": "MEDIUM",
"confidentialityImpact": "NONE",
"integrityImpact": "NONE",
"privilegesRequired": "NONE",
"scope": "UNCHANGED",
"userInteraction": "NONE",
"vectorString": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:L",
"version": "3.1"
},
"format": "CVSS"
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-772",
"description": "Missing Release of Resource after Effective Lifetime",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-09-21T22:04:14.206Z",
"orgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"shortName": "VulnCheck"
},
"references": [
{
"name": "Pull Request #51236",
"tags": [
"issue-tracking",
"patch"
],
"url": "https://github.com/vllm-project/vllm/pull/51236"
},
{
"tags": [
"technical-description"
],
"url": "https://github.com/vllm-project/vllm/blob/v0.29.0/vllm/distributed/kv_transfer/kv_connector/v1/mooncake/mooncake_connector.py#L1234-L1242"
},
{
"tags": [
"product"
],
"url": "https://github.com/vllm-project/vllm"
},
{
"name": "VulnCheck Advisory: vLLM through 0.29.0 Resource Exhaustion via Ownerless Mooncake Transfer Placeholders",
"tags": [
"third-party-advisory"
],
"url": "https://www.vulncheck.com/advisories/vllm-through-0.29.0-resource-exhaustion-via-ownerless-mooncake-transfer-placeholders"
}
],
"title": "vLLM through 0.29.0 Resource Exhaustion via Ownerless Mooncake Transfer Placeholders",
"x_generator": {
"engine": "vulncheck-endgame"
}
}
},
"cveMetadata": {
"assignerOrgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"assignerShortName": "VulnCheck",
"cveId": "CVE-2026-94625",
"datePublished": "2026-09-21T22:04:14.206Z",
"dateReserved": "2026-09-21T21:42:27.917Z",
"dateUpdated": "2026-09-23T15:09:16.926Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-94624 (GCVE-0-2026-94624)
Vulnerability from cvelistv5 – Published: 2026-09-21 22:04 – Updated: 2026-09-22 13:26
VLAI
EPSS
VEX
Title
vLLM through 0.29.0 Denial of Service via Unbounded P2P KV Offloading Sessions
Summary
vLLM through 0.29.0 contains a denial of service vulnerability in P2P KV offloading when OffloadingConnector is configured with TieringOffloadingSpec and a peer-to-peer secondary tier. Attackers can supply arbitrary remote host and port values in kv_transfer_params to create unreachable peer sessions that retain ZeroMQ sockets until the context quota is exhausted, causing an uncaught ZMQError that crashes EngineCore and stops all inference.
Severity
SSVC
Exploitation: none
Automatable: yes
Technical Impact: partial
CISA Coordinator · CISA-ADP (v2.0.3)
Decision recorded 2026-09-22 13:26 UTC
CWE
- CWE-770 - Allocation of Resources Without Limits or Throttling
Assigner
References
5 references
| URL | Tags |
|---|---|
| https://github.com/vllm-project/vllm/pull/51504 | issue-trackingpatch |
| 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://github.com/vllm-project/vllm | product |
| https://www.vulncheck.com/advisories/vllm-through… | third-party-advisory |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| vllm-project | vllm |
Affected:
0 , ≤ 0.29.0
(semver)
cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:* |
Date Public
2026-08-08 00:00
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-94624",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "yes"
},
{
"Technical Impact": "partial"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-09-22T13:26:32.508111Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-09-22T13:26:57.950Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"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": "finder",
"value": "Mingkai Yu"
},
{
"lang": "en",
"type": "finder",
"value": "Jiapeng Li"
},
{
"lang": "en",
"type": "finder",
"value": "Jiajia Liu"
}
],
"datePublic": "2026-08-08T00:00:00.000Z",
"descriptions": [
{
"lang": "en",
"value": "vLLM through 0.29.0 contains a denial of service vulnerability in P2P KV offloading when OffloadingConnector is configured with TieringOffloadingSpec and a peer-to-peer secondary tier. Attackers can supply arbitrary remote host and port values in kv_transfer_params to create unreachable peer sessions that retain ZeroMQ sockets until the context quota is exhausted, causing an uncaught ZMQError that crashes EngineCore and stops all inference."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "LOW",
"attackRequirements": "NONE",
"attackVector": "NETWORK",
"baseScore": 8.7,
"baseSeverity": "HIGH",
"privilegesRequired": "NONE",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "NONE",
"vectorString": "CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:N/VI:N/VA:H/SC:N/SI:N/SA:N",
"version": "4.0",
"vulnAvailabilityImpact": "HIGH",
"vulnConfidentialityImpact": "NONE",
"vulnIntegrityImpact": "NONE"
},
"format": "CVSS"
},
{
"cvssV3_1": {
"attackComplexity": "LOW",
"attackVector": "NETWORK",
"availabilityImpact": "HIGH",
"baseScore": 7.5,
"baseSeverity": "HIGH",
"confidentialityImpact": "NONE",
"integrityImpact": "NONE",
"privilegesRequired": "NONE",
"scope": "UNCHANGED",
"userInteraction": "NONE",
"vectorString": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H",
"version": "3.1"
},
"format": "CVSS"
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-770",
"description": "Allocation of Resources Without Limits or Throttling",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-09-21T22:04:13.172Z",
"orgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"shortName": "VulnCheck"
},
"references": [
{
"name": "Pull Request #51504",
"tags": [
"issue-tracking",
"patch"
],
"url": "https://github.com/vllm-project/vllm/pull/51504"
},
{
"tags": [
"technical-description"
],
"url": "https://github.com/vllm-project/vllm/blob/v0.29.0/vllm/v1/kv_offload/tiering/p2p/control/zmq.py#L240-L260"
},
{
"tags": [
"technical-description"
],
"url": "https://github.com/vllm-project/vllm/blob/v0.29.0/vllm/v1/kv_offload/tiering/p2p/manager.py#L667-L677"
},
{
"tags": [
"product"
],
"url": "https://github.com/vllm-project/vllm"
},
{
"name": "VulnCheck Advisory: vLLM through 0.29.0 Denial of Service via Unbounded P2P KV Offloading Sessions",
"tags": [
"third-party-advisory"
],
"url": "https://www.vulncheck.com/advisories/vllm-through-0.29.0-denial-of-service-via-unbounded-p2p-kv-offloading-sessions"
}
],
"title": "vLLM through 0.29.0 Denial of Service via Unbounded P2P KV Offloading Sessions",
"x_generator": {
"engine": "vulncheck-endgame"
}
}
},
"cveMetadata": {
"assignerOrgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"assignerShortName": "VulnCheck",
"cveId": "CVE-2026-94624",
"datePublished": "2026-09-21T22:04:13.172Z",
"dateReserved": "2026-09-21T21:42:26.965Z",
"dateUpdated": "2026-09-22T13:26:57.950Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-94623 (GCVE-0-2026-94623)
Vulnerability from cvelistv5 – Published: 2026-09-21 22:04 – Updated: 2026-09-22 17:44
VLAI
EPSS
VEX
Title
vLLM through 0.29.0 Denial of Service via NIXL Multi-Prompt Assertion Failure
Summary
vLLM through 0.29.0 contains a denial of service vulnerability in the NIXL connector's prefix caching implementation that fails to properly validate block counts across multi-prompt completion requests in prefill/decode disaggregated deployments. Attackers can trigger an assertion failure in NixlBaseConnectorWorker._apply_prefix_caching by submitting completion requests with multiple prompts of varying lengths, causing the decode worker to terminate and become unavailable until restarted.
Severity
SSVC
Exploitation: none
Automatable: yes
Technical Impact: partial
CISA Coordinator · CISA-ADP (v2.0.3)
Decision recorded 2026-09-22 17:19 UTC
CWE
- CWE-617 - Reachable Assertion
Assigner
References
4 references
| URL | Tags |
|---|---|
| https://github.com/vllm-project/vllm/pull/51505 | issue-trackingpatch |
| https://github.com/vllm-project/vllm/blob/v0.29.0… | technical-description |
| https://github.com/vllm-project/vllm | product |
| https://www.vulncheck.com/advisories/vllm-through… | third-party-advisory |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| vllm-project | vllm |
Affected:
0 , ≤ 0.29.0
(semver)
cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:* |
Date Public
2026-08-08 00:00
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-94623",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "yes"
},
{
"Technical Impact": "partial"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-09-22T17:19:44.934826Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-09-22T17:44:56.345Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"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": "finder",
"value": "Mingkai Yu"
},
{
"lang": "en",
"type": "finder",
"value": "Jiapeng Li"
},
{
"lang": "en",
"type": "finder",
"value": "Jiajia Liu"
}
],
"datePublic": "2026-08-08T00:00:00.000Z",
"descriptions": [
{
"lang": "en",
"value": "vLLM through 0.29.0 contains a denial of service vulnerability in the NIXL connector\u0027s prefix caching implementation that fails to properly validate block counts across multi-prompt completion requests in prefill/decode disaggregated deployments. Attackers can trigger an assertion failure in NixlBaseConnectorWorker._apply_prefix_caching by submitting completion requests with multiple prompts of varying lengths, causing the decode worker to terminate and become unavailable until restarted."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "LOW",
"attackRequirements": "NONE",
"attackVector": "NETWORK",
"baseScore": 8.7,
"baseSeverity": "HIGH",
"privilegesRequired": "NONE",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "NONE",
"vectorString": "CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:N/VI:N/VA:H/SC:N/SI:N/SA:N",
"version": "4.0",
"vulnAvailabilityImpact": "HIGH",
"vulnConfidentialityImpact": "NONE",
"vulnIntegrityImpact": "NONE"
},
"format": "CVSS"
},
{
"cvssV3_1": {
"attackComplexity": "LOW",
"attackVector": "NETWORK",
"availabilityImpact": "HIGH",
"baseScore": 7.5,
"baseSeverity": "HIGH",
"confidentialityImpact": "NONE",
"integrityImpact": "NONE",
"privilegesRequired": "NONE",
"scope": "UNCHANGED",
"userInteraction": "NONE",
"vectorString": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H",
"version": "3.1"
},
"format": "CVSS"
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-617",
"description": "Reachable Assertion",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-09-21T22:04:12.106Z",
"orgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"shortName": "VulnCheck"
},
"references": [
{
"name": "Pull Request #51505",
"tags": [
"issue-tracking",
"patch"
],
"url": "https://github.com/vllm-project/vllm/pull/51505"
},
{
"tags": [
"technical-description"
],
"url": "https://github.com/vllm-project/vllm/blob/v0.29.0/vllm/distributed/kv_transfer/kv_connector/v1/nixl/base_worker.py#L2642-L2646"
},
{
"tags": [
"product"
],
"url": "https://github.com/vllm-project/vllm"
},
{
"name": "VulnCheck Advisory: vLLM through 0.29.0 Denial of Service via NIXL Multi-Prompt Assertion Failure",
"tags": [
"third-party-advisory"
],
"url": "https://www.vulncheck.com/advisories/vllm-through-0.29.0-denial-of-service-via-nixl-multi-prompt-assertion-failure"
}
],
"title": "vLLM through 0.29.0 Denial of Service via NIXL Multi-Prompt Assertion Failure",
"x_generator": {
"engine": "vulncheck-endgame"
}
}
},
"cveMetadata": {
"assignerOrgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"assignerShortName": "VulnCheck",
"cveId": "CVE-2026-94623",
"datePublished": "2026-09-21T22:04:12.106Z",
"dateReserved": "2026-09-21T21:42:25.636Z",
"dateUpdated": "2026-09-22T17:44:56.345Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
CVE-2026-94622 (GCVE-0-2026-94622)
Vulnerability from cvelistv5 – Published: 2026-09-21 22:04 – Updated: 2026-09-22 13:01
VLAI
EPSS
VEX
Title
vLLM through 0.29.0 Denial of Service via Incomplete NIXL KV Transfer Metadata
Summary
vLLM versions through 0.29.0 contain a denial of service vulnerability in the NIXL connector's metadata handling for prefill/decode disaggregated deployments. Attackers can send requests with incomplete kv_transfer_params dictionary entries to trigger an uncaught KeyError in EngineCore scheduling, causing the decode engine to terminate and making all routed requests fail until manual restart.
Severity
SSVC
Exploitation: none
Automatable: yes
Technical Impact: partial
CISA Coordinator · CISA-ADP (v2.0.3)
Decision recorded 2026-09-22 12:59 UTC
CWE
- CWE-248 - Uncaught Exception
Assigner
References
4 references
| URL | Tags |
|---|---|
| https://github.com/vllm-project/vllm/pull/54807 | issue-trackingpatch |
| https://github.com/vllm-project/vllm/blob/v0.29.0… | technical-description |
| https://github.com/vllm-project/vllm | product |
| https://www.vulncheck.com/advisories/vllm-through… | third-party-advisory |
Impacted products
1 product
| Vendor | Product | Version | |
|---|---|---|---|
| vllm-project | vllm |
Affected:
0 , ≤ 0.29.0
(semver)
cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:* |
Date Public
2026-09-01 00:00
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-94622",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "yes"
},
{
"Technical Impact": "partial"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-09-22T12:59:50.196633Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-09-22T13:01:33.609Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"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": "finder",
"value": "Mingkai Yu"
},
{
"lang": "en",
"type": "finder",
"value": "Jiapeng Li"
},
{
"lang": "en",
"type": "finder",
"value": "Jiajia Liu"
}
],
"datePublic": "2026-09-01T00:00:00.000Z",
"descriptions": [
{
"lang": "en",
"value": "vLLM versions through 0.29.0 contain a denial of service vulnerability in the NIXL connector\u0027s metadata handling for prefill/decode disaggregated deployments. Attackers can send requests with incomplete kv_transfer_params dictionary entries to trigger an uncaught KeyError in EngineCore scheduling, causing the decode engine to terminate and making all routed requests fail until manual restart."
}
],
"metrics": [
{
"cvssV4_0": {
"attackComplexity": "LOW",
"attackRequirements": "NONE",
"attackVector": "NETWORK",
"baseScore": 8.7,
"baseSeverity": "HIGH",
"privilegesRequired": "NONE",
"subAvailabilityImpact": "NONE",
"subConfidentialityImpact": "NONE",
"subIntegrityImpact": "NONE",
"userInteraction": "NONE",
"vectorString": "CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:N/VI:N/VA:H/SC:N/SI:N/SA:N",
"version": "4.0",
"vulnAvailabilityImpact": "HIGH",
"vulnConfidentialityImpact": "NONE",
"vulnIntegrityImpact": "NONE"
},
"format": "CVSS"
},
{
"cvssV3_1": {
"attackComplexity": "LOW",
"attackVector": "NETWORK",
"availabilityImpact": "HIGH",
"baseScore": 7.5,
"baseSeverity": "HIGH",
"confidentialityImpact": "NONE",
"integrityImpact": "NONE",
"privilegesRequired": "NONE",
"scope": "UNCHANGED",
"userInteraction": "NONE",
"vectorString": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H",
"version": "3.1"
},
"format": "CVSS"
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-248",
"description": "Uncaught Exception",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-09-21T22:04:10.761Z",
"orgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"shortName": "VulnCheck"
},
"references": [
{
"name": "Pull Request #54807",
"tags": [
"issue-tracking",
"patch"
],
"url": "https://github.com/vllm-project/vllm/pull/54807"
},
{
"tags": [
"technical-description"
],
"url": "https://github.com/vllm-project/vllm/blob/v0.29.0/vllm/distributed/kv_transfer/kv_connector/v1/nixl/metadata.py#L294-L312"
},
{
"tags": [
"product"
],
"url": "https://github.com/vllm-project/vllm"
},
{
"name": "VulnCheck Advisory: vLLM through 0.29.0 Denial of Service via Incomplete NIXL KV Transfer Metadata",
"tags": [
"third-party-advisory"
],
"url": "https://www.vulncheck.com/advisories/vllm-through-0.29.0-denial-of-service-via-incomplete-nixl-kv-transfer-metadata"
}
],
"title": "vLLM through 0.29.0 Denial of Service via Incomplete NIXL KV Transfer Metadata",
"x_generator": {
"engine": "vulncheck-endgame"
}
}
},
"cveMetadata": {
"assignerOrgId": "83251b91-4cc7-4094-a5c7-464a1b83ea10",
"assignerShortName": "VulnCheck",
"cveId": "CVE-2026-94622",
"datePublished": "2026-09-21T22:04:10.761Z",
"dateReserved": "2026-09-21T21:42:24.373Z",
"dateUpdated": "2026-09-22T13:01:33.609Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}