vLLM is an inference and serving engine for large language models (LLMs). From 0.6.1 to before 0.20.0, there is a a Token Injection vulnerability in vLLM’s multimodal processing. Unauthenticated, text-only prompts that spell special tokens are interpreted as control. Image and video placeholder sequences supplied without matching data cause vLLM to index into empty grids during input-position computation, raising an unhandled IndexError and terminating the worker or degrading availability. Multimodal paths that rely on image_grid_thw/video_grid_thw are affected. This vulnerability is fixed in 0.20.0.
{
"cna_assigner": "GitHub_M",
"cwe_ids": [
"CWE-129"
],
"osv_generated_from": "https://github.com/CVEProject/cvelistV5/tree/main/cves/2026/44xxx/CVE-2026-44222.json"
}{
"extracted_events": [
{
"range": [
{
"introduced": "0.6.1"
},
{
"fixed": "0.20.0"
}
],
"source": "AFFECTED_FIELD",
"strategy": "StringRangeExpression"
},
{
"cpe": "cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:*",
"range": [
{
"introduced": "0.6.1"
},
{
"fixed": "0.20.0"
}
],
"source": "CPE_RANGE"
}
]
}