Home › Vulnerabilities

vLLM Flaw CVE-2026-115915: Update Now

Published October 29, 2026 · Last updated October 29, 2026

CVE Critical Patch

The short version: CVE-2026-115915 is a critical-severity security flaw in vLLM. The National Vulnerability Database (NVD) published it on October 26, 2026, and the score it lists is 9.8 out of 10. When FraudMeNot checked on October 29, 2026, CISA had not added CVE-2026-115915 to its Known Exploited Vulnerabilities catalog, so there was no public U.S. government confirmation of attacks yet. A fix is available, which is why this page focuses on who should update and how.

What the official record says about CVE-2026-115915

The record was filed by github.com. In its own words: "vLLM is an inference and serving engine for large language models. Prior to 0.32.0, ImageMediaIO passes attacker-controlled image bytes to Pillow through Image.open(BytesIO(data)) without restricting decoder selection, and the MIME label does not constrain the selected format. Image-capable request paths through parse_image() and fetch_image() can therefore dispatch an EPS payload to host components. A host with Ghostscript affected by CVE-2024-29510 can execute commands with the vLLM process's privileges, while environments using Pillow 12.0.0 through 12.2.0 can trigger CVE-2026-59203 and exhaust media-loading workers. Deployments without the applicable vulnerable dependency are not affected by that corresponding impact. This issue is fixed in version 0.32.0."

Why vLLM admins should care

Severity scores use the Common Vulnerability Scoring System (CVSS). the address listed on nist.gov rated CVE-2026-115915 9.8 (Critical) under CVSS 3.1. Reading that rating in everyday terms: it can be reached over a network, including the internet; an attacker does not need an account; no one has to click or open anything. If it is abused, the rated damage to keeping data private is high, to keeping data unaltered is high, and to keeping vLLM running is high. Scoring string: CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H.

Scores can differ between the vendor and NVD because each judges the worst realistic case differently. For this flaw, the address listed on github.com gives it 8.1 (High, CVSS 3.1).

Affected vLLM releases

According to the record, these releases are vulnerable:

  • version < 0.32.0

The type of bug behind CVE-2026-115915

NVD classifies it using MITRE's Common Weakness Enumeration:

  • CWE-20, Improper Input Validation. MITRE describes this class of bug as follows: "The product receives input or data, but it does not validate or incorrectly validates that the input has the properties that are required to process the data safely and correctly."
  • CWE-183, Permissive List of Allowed Inputs. MITRE describes this class of bug as follows: "The product implements a protection mechanism that relies on a list of inputs (or properties of inputs) that are explicitly allowed by policy because the inputs are assumed to be safe, but the list is too permissive - that is, it allows an input that is unsafe, leading to resultant weaknesses."

Vendor remediation for CVE-2026-115915

The vendor advisory remediation for operators of vLLM, quoted for this page, is: "Upgrade vLLM to version 0.32.0 or later (GitHub advisory GHSA-h3rc-j82p-2c3h). Until you can, do not accept image input from untrusted users, and make sure Ghostscript on the host is patched and Pillow is outside the 12.0.0 to 12.2.0 range."

What to do about it

  • Update now: Upgrade vLLM to version 0.32.0 or later (GitHub advisory GHSA-h3rc-j82p-2c3h). Until you can, do not accept image input from untrusted users, and make sure Ghostscript on the host is patched and Pillow is outside the 12.0.0 to 12.2.0 range.
  • Who should act: IT teams, managed service providers, and developers who run or ship vLLM. Check every place it runs, including servers, build systems, containers, and older machines nobody watches closely.
  • Read the official advisory from vLLM (linked under Sources) for exact fixed builds and any temporary workaround if you cannot patch today.
  • After patching, confirm the new version is actually running and review logs from October 26, 2026 onward for anything unusual.
  • Home users: you do not need to buy or download anything because of CVE-2026-115915. A caller, text, or pop-up claiming your computer is "infected" with this CVE and asking for money or remote access is a scam.

CVE-2026-115915 record details from CISA and NVD

  • Attack complexity for CVE-2026-115915: low
  • CVE-2026-115915 CVSS scope: unchanged
  • NVD sub-scores for CVE-2026-115915: exploitability 3.9, impact 5.9.
  • Who mapped the weakness type: the address listed on github.com (Secondary)
  • NVD product identifiers (CPE): cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:*
  • References in the NVD record: 4 (3 tagged Patch, 1 tagged Release Notes, 1 tagged Vendor Advisory).
  • CISA SSVC decision points for CVE-2026-115915 (from NVD): exploitation=none; automatable=no; technicalImpact=total (role: CISA Coordinator, SSVC 2.0.3).

Dates to know

  • October 26, 2026: NVD publishes CVE-2026-115915.
  • October 29, 2026: FraudMeNot checks the CISA KEV catalog; CVE-2026-115915 is not listed.
  • October 28, 2026: most recent NVD update (record status: Analyzed).

Official references

Links listed in the NVD record (labels in brackets come from NVD):

Sources

General education only, not security advice. Facts checked against the sources above on the date shown.