CVE-2026-114055

validated

GHSA-3cgq-c484-rx5j ↗EUVD-2026-107006 ↗

check availablenot proven exploitable

cve.orgpublic pockev

vllm-project vllm >= 0.7.3, < 0.31.0

fixed in 0.31.0

ETK-2026-2282, permanent identifier

cvss7.5cna vendorCVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H

+4h13mTTV

35mAI analysis

Check if you are vulnerable

timeline

0 = trigger signal

trigger signal (cve.org) +0hcheck available+4h13magent stopped

+1h

+2h

+3h

description

vLLM versions from 0.7.3 through versions earlier than 0.31.0 are affected when multimodal request processing accepts client-supplied processor options. A remote client may cause excessive memory use and interrupt the shared service by submitting requests with oversized multimodal processing settings.

mitigation

Upgrade vLLM to version 0.31.0 or later. Until upgraded, restrict access to the multimodal API and reject or tightly limit client-controlled multimodal processor options.

fixed in 0.31.0

check

available

exploit

automation

deterministic

auth required

EPSS

–

agent run time

35m

cve.org record, t = 0

2026-10-21 00:36Z

Δ to public poc

–

Δ to kev

–

outcome

Check only, differential: the check fires on the vulnerable build and is silent on the patched one.

source signals

Public signals that triggered or updated the pipeline. These are inputs for the deltas above.

2026-10-21 00:36Z
ghsa: trigger signal, started the run
https://github.com/advisories/GHSA-3cgq-c484-rx5j
2026-10-21 00:36Z
cve.org record for CVE-2026-114055
cve_org · https://www.cve.org/CVERecord?id=CVE-2026-114055

this entry in the feed

GET /api/v1/feed/CVE-2026-114055

{
  "etk_id": "ETK-2026-2282",
  "cve_id": "CVE-2026-114055",
  "display_id": "CVE-2026-114055",
  "ghsa_id": "GHSA-3cgq-c484-rx5j",
  "euvd_id": "EUVD-2026-107006",
  "status": "published",
  "anchor_state": "anchored",
  "title": "Uncontrolled resource consumption in vLLM multimodal processing",
  "summary": "vLLM versions from 0.7.3 through versions earlier than 0.31.0 are affected when multimodal request processing accepts client-supplied processor options. A remote client may cause excessive memory use and interrupt the shared service by submitting requests with oversized multimodal processing settings.",
  "mitigation": "Upgrade vLLM to version 0.31.0 or later. Until upgraded, restrict access to the multimodal API and reject or tightly limit client-controlled multimodal processor options.",
  "vendor": "vllm-project",
  "product": "vllm",
  "affected": [
    {
      "affected_notes": [],
      "affected_ranges": [
        ">= 0.7.3, < 0.31.0"
      ],
      "ecosystem": null,
      "fixed_versions": [
        "0.31.0"
      ],
      "package_name": "vllm",
      "purl": null
    }
  ],
  "cwes": null,
  "cvss_score": 7.5,
  "cvss_src": "cna_vendor",
  "cvss_vector": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H",
  "epss_score": null,
  "epss_percentile": null,
  "epss_asof": null,
  "automation": "deterministic",
  "check": true,
  "check_confidence_basis": "differential",
  "exploit": false,
  "pipeline_minutes": 35.4534951,
  "t0_at": "2026-10-21T00:36:31.227000+00:00",
  "signals": {
    "check": {
      "at": "2026-10-21T04:49:26.847833+00:00",
      "days": 0.1756437596412037,
      "is_lower_bound": false,
      "measured_at": "2026-10-21T00:36:31.227000+00:00",
      "url": null
    },
    "trigger_signal": {
      "at": "2026-10-21T00:36:31.227000+00:00",
      "days": 0,
      "is_lower_bound": false,
      "measured_at": "2026-10-21T00:36:31.227000+00:00",
      "url": "https://github.com/advisories/GHSA-3cgq-c484-rx5j"
    }
  }
}

published 2026-10-21 04:13Z
last change 2026-10-21 05:00Z

Need the technical details behind an entry?

Every entry here shows what we attempted and when. The technical details, the generated exploit that fired it and the evidence behind the verdict are not public. Tell us about your use case and we’ll get back to you about extended API access.