UTC
High

MOSS-Audio request-controlled processor cache permits remote memory exhaustion

CVE-2026-114052 · GHSA-823j-m4cj-hmjf · published 2026-10-06 · updated 2026-10-23

Severity
High
CVSS score
7.5
Chance of exploitation (EPSS)
Not scored
Fixed in
0.31.0
  • Exploitation No known exploitation
  • Automatable Yes
  • Technical impact Partial
  • CISA SSVC
What it is

An attacker who can reach a vLLM HTTP server serving a MOSS-Audio model can cause persistent attacker-controlled memory allocations in the frontend API process. Repeated unique configurations cause linear retained-memory growth. Oversized mel_config values can also cause a very large allocation from a single request. Continued exploitation can terminate the process through out-of-memory exhaustion, resulting in denial of service. The issue is limited to deployments serving a MOSS-Audio model. A reverse proxy that authenticates or rate-limits / tokenize can reduce exposure, but vLLM’s own --api-key setting does not protect this endpoint.

What to do

Upgrade vllm to 0.31.0 or later. If you start it from an MCP config with npx or uvx, pin that version so an old cached copy is not used.

Affected
PackageRegistryAffectedFixed in
vllmPyPI> 0.24.0, < 0.31.00.31.0

The impact, patch and workaround text comes from the advisory, shortened (published by vLLM). Proof-of-concept and reproduction sections and code are left out; the original advisory has them.

Sources: GitHub Advisory Database, reviewed and malware advisories (CC BY 4.0, https://github.com/advisories); advisories published by Cursor (https://github.com/cursor/cursor/security/advisories); advisories published by OpenClaw (https://github.com/openclaw/openclaw/security/advisories); advisories published by vLLM (https://github.com/vllm-project/vllm/security/advisories); advisories published by Open WebUI (https://github.com/open-webui/open-webui/security/advisories); advisories published by n8n (https://github.com/n8n-io/n8n/security/advisories); advisories published by Flowise (https://github.com/FlowiseAI/Flowise/security/advisories); advisories published by Langflow (https://github.com/langflow-ai/langflow/security/advisories); advisories published by the MCP Python SDK project (https://github.com/modelcontextprotocol/python-sdk/security/advisories); advisories published by MLflow (https://github.com/mlflow/mlflow/security/advisories); CISA Known Exploited Vulnerabilities catalog (CC0). Data as of 2026-11-22. ShieldMap describes how each problem works and how to fix it; it does not publish exploit code. Report a correction to [email protected].