CVE intelligence and bounded remediation

CVE-2025-62164 — vLLM is an inference and serving engine for large language models (LLMs)

High CVSS 8.8

vLLM is an inference and serving engine for large language models (LLMs). From versions 0.10.2 to before 0.11.1, a memory corruption vulnerability could lead to a crash (denial-of-service) and potentially remote code execution (RCE), exists in the Completions API endpoint. When processing user-supplied prompt embeddings, the endpoint loads serialized tensors using torch.load() without sufficient validation. Due to a change introduced in PyTorch 2.8.0, sparse tensor integrity checks are disabled by default. As a result, maliciously crafted tensors can bypass internal bounds checks and trigger an out-of-bounds memory write during the call to to_dense(). This memory corruption can crash vLLM and potentially lead to code execution on the server hosting vLLM. This issue has been patched in version 0.11.1.

Severity
High
CVSS
8.8 (3.1)
Published
2025-11-21
CISA KEV
Not currently listed
Ecosystem
software/application
Weaknesses
CWE-20, CWE-123, CWE-502, CWE-787

Affected products

  • vllm / vllm
  • vllm / vllm / 0.11.1

Showing 2 representative product identities from 3 source matches. Confirm exact affected versions with the linked vendor and NVD evidence.

Matched remediation archetype

Unsafe deserialization and object reconstruction

This catalog composition supplies bounded fallback guidance. Explicitly reviewed curated workflows load with the complete record below.

Check exposure

  • Inventory serialization formats accepted from requests, queues, caches, files, cookies, and cross-service messages.
  • Trace whether untrusted input can select classes, types, callbacks, constructors, or object hooks during decoding.
  • Identify signing, schema validation, trust-boundary, and compatibility settings for each decoder.

Remediate safely

  • Replace native object deserialization with a data-only format and explicit schema validation.
  • If replacement is not immediate, use a safe decoder with a minimal type allowlist and disable polymorphic or executable hooks.
  • Update the affected library and add inert tests for unknown types, extra fields, malformed nesting, and unsigned data.

Authoritative sources

Complete CVE record and remediation plan

The detailed catalog view below loads this exact record, its source evidence, and the full seven-phase agentic change plan.