vLLM before 0.29.0 Resource Exhaustion via Unbounded Media Materialization

Summary

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.

Timeline

1. Sep 26, 2026, 1:23 PM UTCCVE published
2. Sep 30, 2026, 2:49 PM UTCCISA SSVC assessment

   Exploitation: poc · Automatable: no · Technical impact: partial

3. Oct 1, 2026, 11:06 AM UTCRecord last updated

Affected versions

• vllm-project vllm

  fixed in 0.29.0

• vllmPyPI

  fixed in 0.30.0

References 3