Advisories

vLLM before 0.29.0 Resource Exhaustion via Unbounded Media Materialization

Go Back
severity
high
date
Affecting
  • vllm >= 0, < 0.29.0

CWE
  • CWE-400 Uncontrolled Resource Consumption
CVSS
7.1
CVSS V4 Vector
CVSS:4.0/AV:N/AC:L/AT:N/PR:L/UI:N/VC:N/VI:N/VA:H/SC:N/SI:N/SA:N
Credit
KernelClint, jperezdealgaba
Description
vLLM through 0.29.0 fetches and fully materializes remote or inline media before enforcing its documented media controls (the VLLM_MAX_AUDIO_CLIP_FILESIZE_MB compressed-audio size cap, default 25 MB, and the per-modality --limit-mm-per-prompt item limits). Across four ingress paths — the shared media-acquisition layer (HTTPConnection.get_bytes()/async_get_bytes()), the chat completions audio_url/base64 path, the batch speech runner, and the Rust frontend POST /tokenize route — the server reads the entire HTTP response body, base64-decodes the inline payload, or spawns one fetch/decode task per media part, and only then applies the limit (or, on some paths, never applies it). A remote attacker can therefore cause the API server or batch-runner process to allocate memory and consume outbound bandwidth proportional to an attacker-chosen body size or media item count before the request is rejected, resulting in pre-inference memory and bandwidth exhaustion (denial of service). The chat and batch surfaces require an API key when one is configured; the Rust frontend /tokenize route is unauthenticated by design. There is no code execution or data disclosure impact.

Ready to get Started?

Explore VulnCheck, a next-generation Cyber Threat Intelligence platform, which provides exploit and vulnerability intelligence to help you prioritize and remediate vulnerabilities that matter.
  • Vulnerability Prioritization
    Prioritize vulnerabilities that matter based on the threat landscape and defer vulnerabilities that don't.
  • Early Warning System
    Real-time alerting of changes in the vulnerability landscape so that you can take action before the attacks start.