CVE-2026-93436Published Sep 17, 2026 · Sep 17Updated Oct 28
vLLM through 0.29.0 Memory Exhaustion via Rejected Requests
vllm-project vllm
High8.7Fix available
Affected versions
| Product | Affected | Fixed in |
|---|---|---|
| vllm PyPI |
|
|
Original description
vLLM through 0.29.0 fails to properly clean up decode-side metadata for rejected inference requests in prefill/decode disaggregated deployments. Remote attackers can submit requests with max_tokens=0 to exhaust decode-worker memory without bound until the worker restarts.
From the CVE record · VulnCheck