Skip to main content

Vllm CVE-2025-47277

CRITICAL
Deserialization of Untrusted Data (CWE-502)
2025-05-20 security-advisories@github.com
Critical
Disputed · 9.8 NVD
Share

Severity by source

Sources disagree (Medium–Critical)
GitHub Advisory PRIMARY
9.8 CRITICAL
AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H
Red Hat
7.5 MEDIUM
qualitative

vuln.today treats the vendor’s rating as authoritative. A higher third-party CVSS (e.g. CISA-ADP) is shown for transparency but does not drive the headline severity.

CVSS VectorGitHub Advisory

CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H
Attack Vector
Network
Attack Complexity
Low
Privileges Required
None
User Interaction
None
Scope
Unchanged
Confidentiality
High
Integrity
High
Availability
High

Lifecycle Timeline

4
Analysis Generated
Mar 28, 2026 - 18:42 vuln.today
Patch released
Mar 28, 2026 - 18:42 nvd
Patch available
PoC Detected
Aug 13, 2025 - 16:35 vuln.today
Public exploit code
CVE Published
May 20, 2025 - 18:15 nvd
CRITICAL 9.8

DescriptionGitHub Advisory

vLLM, an inference and serving engine for large language models (LLMs), has an issue in versions 0.6.5 through 0.8.4 that ONLY impacts environments using the PyNcclPipe KV cache transfer integration with the V0 engine. No other configurations are affected. vLLM supports the use of the PyNcclPipe class to establish a peer-to-peer communication domain for data transmission between distributed nodes. The GPU-side KV-Cache transmission is implemented through the PyNcclCommunicator class, while CPU-side control message passing is handled via the send_obj and recv_obj methods on the CPU side.​ The intention was that this interface should only be exposed to a private network using the IP address specified by the --kv-ip CLI parameter. The vLLM documentation covers how this must be limited to a secured network. The default and intentional behavior from PyTorch is that the TCPStore interface listens on ALL interfaces, regardless of what IP address is provided. The IP address given was only used as a client-side address to use. vLLM was fixed to use a workaround to force the TCPStore instance to bind its socket to a specified private interface. As of version 0.8.5, vLLM limits the TCPStore socket to the private interface as configured.

AnalysisAI

vLLM, an inference and serving engine for large language models (LLMs), has an issue in versions 0.6.5 through 0.8.4 that ONLY impacts environments using the PyNcclPipe KV cache transfer. Rated critical severity (CVSS 9.8), this vulnerability is remotely exploitable, no authentication required, low attack complexity. Public exploit code available.

Technical ContextAI

This vulnerability is classified as Deserialization of Untrusted Data (CWE-502), which allows attackers to execute arbitrary code through malicious serialized objects. vLLM, an inference and serving engine for large language models (LLMs), has an issue in versions 0.6.5 through 0.8.4 that ONLY impacts environments using the PyNcclPipe KV cache transfer integration with the V0 engine. No other configurations are affected. vLLM supports the use of the PyNcclPipe class to establish a peer-to-peer communication domain for data transmission between distributed nodes. The GPU-side KV-Cache transmission is implemented through the PyNcclCommunicator class, while CPU-side control message passing is handled via the send_obj and recv_obj methods on the CPU side.​ The intention was that this interface should only be exposed to a private network using the IP address specified by the --kv-ip CLI parameter. The vLLM documentation covers how this must be limited to a secured network. The default and intentional behavior from PyTorch is that the TCPStore interface listens on ALL interfaces, regardless of what IP address is provided. The IP address given was only used as a client-side address to use. vLLM was fixed to use a workaround to force the TCPStore instance to bind its socket to a specified private interface. As of version 0.8.5, vLLM limits the TCPStore socket to the private interface as configured. Affected products include: Vllm. Version information: through 0.8.4.

RemediationAI

A vendor patch is available. Apply the latest security update as soon as possible. Avoid deserializing untrusted data. Use safe serialization formats (JSON). Implement integrity checks and type allowlists.

More in Vllm

View all
CVE-2025-32444 CRITICAL POC
10.0 Apr 30

vLLM is a high-throughput and memory-efficient inference and serving engine for LLMs. Rated critical severity (CVSS 10.0

CVE-2024-11041 CRITICAL POC
9.8 Mar 20

vllm-project vllm version v0.6.2 contains a vulnerability in the MessageQueue.dequeue() API function. Rated critical sev

CVE-2026-22778 CRITICAL POC
9.8 Feb 02

Information exposure in vLLM inference engine versions 0.8.3 to before 0.14.1. Invalid image requests to the multimodal

CVE-2025-30202 HIGH POC
7.5 Apr 30

vLLM is a high-throughput and memory-efficient inference and serving engine for LLMs. Rated high severity (CVSS 7.5), th

CVE-2026-24779 HIGH POC
7.1 Jan 27

vLLM before version 0.14.1 contains a server-side request forgery vulnerability in the MediaConnector class where incons

CVE-2025-46560 MEDIUM POC
6.5 Apr 30

vLLM is a high-throughput and memory-efficient inference and serving engine for LLMs. Rated medium severity (CVSS 6.5),

CVE-2026-22773 MEDIUM POC
6.5 Jan 10

Vllm versions up to 0.12.0 is affected by allocation of resources without limits or throttling (CVSS 6.5).

CVE-2026-22807 CRITICAL
9.8 Jan 21

Remote code execution in vLLM 0.10.1 through 0.13.x lets an attacker who controls the model repository or path run arbit

CVE-2026-25960 CRITICAL
9.8 Mar 09

Server-Side Request Forgery in vLLM's multimodal MediaConnector allows remote attackers to coerce the inference server i

CVE-2026-9540 MEDIUM POC
5.5 May 26

Denial of service in vllm 0.19.0's OpenAI-compatible serving path allows remote unauthenticated attackers to exhaust sch

CVE-2025-29783 CRITICAL
9.0 Mar 19

vLLM is a high-throughput and memory-efficient inference and serving engine for LLMs. Rated critical severity (CVSS 9.0)

CVE-2026-54232 HIGH
8.8 Jun 22

Remote code execution in vLLM versions prior to 0.22.1 allows attackers to backdoor production LLM inference deployments

Vendor StatusVendor

Share

CVE-2025-47277 vulnerability details – vuln.today

This site uses cookies essential for authentication and security. No tracking or analytics cookies are used. Privacy Policy