CVE-2026-93592

vLLM versions before 0.28.0 fail to validate the lower bound of token IDs in the /v1/embeddings and /pooling endpoints, allowing unauthenticated attackers to crash the engine by submitting negative token IDs. A single request with a negative token ID triggers a CUDA device-side assertion that poisons the GPU context, causing all subsequent requests to fail until the process restarts.

  • Published Sep 18, 2026
  • CVSS 8.7 high
  • 0.5% chance of exploitation in the next 30 days (EPSS)
  • Public exploit code is available
  • A fix is available

Affected software

CVE-2026-93592 at the National Vulnerability Database