vLLM from 0.6.0 before 0.33.0 lacks a request-wide media budget, allowing remote inference clients to exhaust frontend resources by submitting many remote image URLs in one request. Attackers can send a compact request to /v1/messages or /v1/responses that forces downloading and decoding of every image before prompt-length admission, exhausting memory, CPU, and connections.
{
"cna_assigner": "VulnCheck",
"cwe_ids": [
"CWE-400"
],
"osv_generated_from": "https://github.com/CVEProject/cvelistV5/tree/main/cves/2026/130xxx/CVE-2026-130686.json"
}