Read the docs
TI-2026-68364 Aliases CVE-2026-105753GHSA-ph3r-5jfg-f84f

vLLM multimodal cache desync triggers reachable assertion crash

Priority LOW 29CWE-617Published · Updated

Key facts

Attack vector
Network
Privileges required
Low
User interaction
None
Attack complexity
Low
Impact type
Denial of service
Confidentiality · integrity · availability
None · None · High
Fix status
Partial Red Hat: 3 without a fix

Attack profile from CVSS 3.1 · CNA. Fix status from CNA, GitHub and Red Hat. Impact type from CWE-617.

Summary

vLLM before 0.28.0 has a cache desync in its mirrored multimodal IPC caches. A request rejected during admission leaves a media hash in the frontend sender cache that the engine receiver cache never stores, so a later request reusing that hash hits a reachable assertion (CWE-617) in the engine core. A remote attacker with low privileges can crash the shared service, causing an availability failure.

Written with Claude Haiku 5.5 from the sources below · 2026-10-29

Description from the CNA

vLLM is an inference and serving engine for large language models. Prior to 0.28.0, the default mirrored multimodal LRU cache can commit a media hash in the frontend sender cache during multimodal rendering and before engine admission, while the engine receiver cache never receives the payload if that request is rejected. A later request reusing the same media hash causes MultiModalProcessorSenderCache to send no payload and MultiModalReceiverCache to reach an assertion with the message "Expected a cached item," producing a shared-service availability failure. This issue is fixed in version 0.28.0.

Why it matters now

Exploitation None reported CISA SSVC · not in CISA KEV
EPSS 0.00427 37th percentileScored 2026-11-07
Fix Partial vllm (vllm-project): 0.28.0vllm (PyPI): 0.28.0Red Hat: 3 without a fix

SSVC decision points

CISA ADP · SSVC 2.0.3

Exploitation
None
Automatable
No
Technical impact
Partial

The decision (Track, Attend, Act) also depends on your mission and well-being impact; see CISA's SSVC guide.

Affected products

VendorProductAffected versionsFixed inStatusSource
vllm-projectvllmbefore 0.28.00.28.0Patch availableCNA
—vllm (PyPI)before 0.28.00.28.0Patch availableGHSA-ph3r-5jfg-f84f
  • vllm Vendorvllm-project Affected before 0.28.0 Fixed in 0.28.0 Patch available CNA
  • vllm (PyPI) Vendor— Affected before 0.28.0 Fixed in 0.28.0 Patch available GHSA-ph3r-5jfg-f84f

CPE configurations NVD · 1

CPEVersionsRoleVendor
cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:*before 0.28.0VulnerablevLLM
  • cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:* Versions before 0.28.0 Vulnerable ·vLLM

Remediation

Upgrade the PyPI vllm package to 0.28.0. Red Hat has no fix plan available for Red Hat AI Inference Server, Red Hat Enterprise Linux AI (RHEL AI) 3 and Red Hat OpenShift AI (RHOAI).

Written with Claude Haiku 5.5 from the sources below · 2026-10-29

  • Fix · CNA vllm (vllm-project) Fixed in 0.28.0
  • Fix · GHSA-ph3r-5jfg-f84f vllm (PyPI) Fixed in 0.28.0 Advisory →

Vendor assessments

Red Hat

Red Hat rating: Moderate. Red Hat's page for this CVE →

ProductStatusPlan or fix
Red Hat AI Inference ServerAffectedNo fix yet: Fix deferred
Red Hat Enterprise Linux AI (RHEL AI) 3AffectedNo fix yet: Fix deferred
Red Hat OpenShift AI (RHOAI)AffectedNo fix yet: Fix deferred

Timeline

  1. 2026-10-06 Threatist record created from CVE-2026-105753
  2. 2026-10-06 Alias added: GHSA-ph3r-5jfg-f84f
  3. 2026-10-06 Fix available: PyPI/[email protected] · GHSA-ph3r-5jfg-f84f
  4. 2026-10-06 CVSS score added · GHSA-ph3r-5jfg-f84f
  5. 2026-10-06 Weakness changed · GHSA-ph3r-5jfg-f84f
  6. 2026-10-06 6 new references · GHSA-ph3r-5jfg-f84f
  7. 2026-10-06 Affected versions changed · Red Hat
  8. 2026-10-06 CVSS score added · Red Hat
  9. 2026-10-06 CISA SSVC decision points changed
  10. 2026-10-06 Priority changed: MEDIUM → LOW

References

Patch 3

Advisory 1

Other 2

  • TI-2024-10229 Remote Code Execution in vllm-project/vllm CVSS 3.0CRITICAL 9.8
  • TI-2024-08404 Remote Code Execution in vllm-project/vllm CVSS 3.1CRITICAL 9.8
  • TI-2026-36017 vLLM: OpenAI auth bypass CVSS 3.1CRITICAL 9.1
  • TI-2026-63056 vLLM before 0.28.0 Denial of Service via negative token ID CVSS 4.0HIGH 8.7
  • TI-2026-04308 Hardcoded trust_remote_code=True in vllm-project/vllm Bypasses User Security Control CVSS 3.0HIGH 8.8
  • TI-2026-39137 vLLM: Remote DoS in vLLM via Invalid Recovered Token Reinjection CVSS 3.1HIGH 7.5
See all 124 →

Sources

This product uses the NVD API but is not endorsed or certified by the NVD. Sources and licences