{
 "id": "AIR-2026-0432",
 "slug": "cve-2026-22773-vllm-denial-of-service-via-idefics3-vision-mo",
 "title": "CVE-2026-22773: vLLM Denial of Service via Idefics3 Vision Model Payload",
 "summary": "An unhandled runtime error in vLLM occurs when processing a crafted 1x1 pixel image payload for Idefics3 models. This tensor dimension mismatch causes a complete server crash.",
 "type": "attack_on_ai",
 "categories": [
  "infra_vuln"
 ],
 "severity": 2,
 "targets": [
  "framework_or_tooling"
 ],
 "products": [
  "vLLM"
 ],
 "incident_date": "2026-01-13",
 "incident_date_precision": "day",
 "first_reported": "2026-01-10",
 "sources": [
  {
   "url": "https://nvd.nist.gov/vuln/detail/CVE-2026-22773",
   "publisher": "NVD",
   "title": "CVE-2026-22773: vLLM is an inference and serving engine for large language models (LLMs).",
   "published_at": "2026-01-10"
  },
  {
   "url": "https://github.com/advisories/GHSA-grg2-63fw-f2qr",
   "publisher": "GitHub Advisory Database",
   "title": "CVE-2026-22773: vLLM is vulnerable to DoS in Idefics3 vision models via image payload with ambiguous dimensions",
   "published_at": "2026-01-13"
  }
 ],
 "source_count": 2,
 "confidence": 0.95,
 "status": "published",
 "origin": "seed_ghsa",
 "tags": [
  "CVE-2026-22773",
  "dos"
 ],
 "created_at": "2026-10-18T07:00:14.101Z",
 "updated_at": "2026-10-18T10:04:21.885Z"
}