vLLM phi4mm: Quadratic Time Complexity in Input Token Processing leads to denial of service

published 2025-04-29 18 mo ago modified 2026-08-07 3 mo ago also EUVD-2025-12671 , GHSA-vc6m-hm49-g9qg , PYSEC-2026-2022

cvss: 6.5; version 3.1, CNA
cisa kev: Not listed; not in CISA's catalog
epss: 0.5%; EPSS 0.5% chance of exploitation in 30 days, percentile 45%
public exploits: 0; none known

vLLM is a high-throughput and memory-efficient inference and serving engine for LLMs. Versions starting from 0.8.0 and prior to 0.8.5 are affected by a critical performance vulnerability in the input preprocessing logic of the multimodal tokenizer. The code dynamically replaces placeholder tokens (e.g., <|audio_|>, <|image_|>) with repeated tokens based on precomputed lengths. Due to inefficient list concatenation operations, the algorithm exhibits quadratic time complexity (O(n²)), allowing malicious actors to trigger resource exhaustion via specially crafted inputs. This issue has been patched in version 0.8.5.

Patch / fix

fixed in 3

pypi vllm 0.8.5
vllm-project vllm 0.8.5
vllm vllm 0.8.5

Vendor advisories

What the vendors themselves say about their products: affected, fixed or not affected, and what to do.

red hat product security vex CVE-2025-46560 updated 2026-06-27 cvss 6.5

vllm: vLLM phi4mm: Quadratic Time Complexity in Input Token Processing leads to denial of service

Red Hat Enterprise Linux AI (RHEL AI) rhelai1/ui-rhel9, rhelai1/bootc-amd-rhel9, rhelai1/bootc-intel-rhel9, rhelai1/bootc-nvidia-rhel9, rhelai1/bootc-azure-amd-rhel9, rhelai1/instructlab-amd-rhel9, +6 more
not affected: vulnerable code not present

Red Hat CSAF VEX © Red Hat, Inc., from https://security.access.redhat.com/data/csaf/v2/vex/; changed: reduced to product statuses and remediations. CC BY 4.0 · csaf

CISA SSVC

Exploitation: Proof of concept is public
Automatable: no
Technical impact: partial
Assessed: 18 mo ago (2025-04-30)

Exploit prediction

0.53% chance of exploitation in the next 30 days, higher than 45% of all scored CVEs (FIRST EPSS, 2026-11-08).

Exploits and checks

github.com https://github.com/vllm-project/vllm/security/advisories/GHSA-vc6m-hm49-g9qg referenced PoC, tagged "exploit" by a source, unverified

CVSS scores

assigned by	version	score	vector
CNA	3.1	6.5	CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H
GitHub	3.1	6.5	CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H
NVD	3.1	7.5	CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H
OSV	3.1	6.5	CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H

Affected packages

ecosystem	package	affected	fixed in
pypi	vllm	>=0.8.0, <0.8.5	0.8.5

Affected products

vllm-project vllm

Classification

Weakness

CWE-1333 Inefficient Regular Expression Complexity

References

github.com /vllm-project/vllm/security/advisories/GHSA-vc6m-hm49-g9qg
exploit, web, vendor-advisory
github.com /vllm-project/vllm/blob/8cac35ba435906fb7eb07e44fe1a8c26e8744f4e/vllm/model_executor/models/phi4mm.py#L1182-L1
web, product
nvd.nist.gov /vuln/detail/CVE-2025-46560
advisory
github.com /advisories/GHSA-vc6m-hm49-g9qg
advisory
github.com /pypa/advisory-database/tree/main/vulns/vllm/PYSEC-2026-2022.yaml
web
github.com /vllm-project/vllm
package
pypi.org /project/vllm
web, package

Sources

Fencer Vulns keeps one entry per vulnerability: the CVE id wins, then GHSA, and records that alias each other are merged, except that two CVEs always keep separate entries. The same data is available as JSON.

CPEs and fixed versions use Anchore's correction of NVD's data: Added CPE configurations because not yet analyzed by NVD.

Summary