A high-throughput and memory-efficient inference and serving engine for LLMs
[email protected] is safe to use (health: 70/100)
| Severity | ID | Summary | Fixed in |
|---|---|---|---|
| medium | CVE-2026-53923 | vLLM: GGUF dequantize kernel int truncation exposes uninitialized GPU memory in multi-tenant serving | — |
| medium | CVE-2026-54233 | vLLM: OOM Denial of Service via Audio Decompression Bomb | — |
| medium | CVE-2026-54235 | vLLM: temperature=NaN and temperature=Infinity bypass validation and propagate to GPU kernels | — |
| medium | CVE-2026-12491 | vLLM: image EXIF Rotation & PNG tRNS Transparency Not Normalized, Causing Mismatch Between Model Input and Expectations | — |
| medium | CVE-2026-54236 | vLLM: incomplete CVE-2026-22778 fix leaks PIL repr addresses via Anthropic router | — |
Get this data programmatically — free, no authentication.
curl https://depscope.dev/api/check/pypi/vllmLast updated · 2026-06-15T05:11:49.394364Z