vLLM is an inference and serving engine for large language models. Prior to 0.32.0, an authenticated caller can place forged shared-memory address and monotonic_id values in public multimodal features.kwargs_data. In scale-out deployments with shared-memory multimodal caching enabled, workers trust those values and pass attacker-selected bytes to pickle.loads(), allowing arbitrary Python execution even when insecure serialization is disabled. This issue is fixed in version 0.32.0.
{
"cna_assigner": "GitHub_M",
"cwe_ids": [
"CWE-502"
],
"osv_generated_from": "https://github.com/CVEProject/cvelistV5/tree/main/cves/2026/115xxx/CVE-2026-115912.json"
}{
"extracted_events": [
{
"cpe": "cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:*",
"range": [
{
"introduced": "0"
},
{
"fixed": "0.32.0"
}
],
"source": "CPE_RANGE"
},
{
"range": [
{
"fixed": "3d5625cc3c2474994107047d7a6ecb8e4008f9b0"
}
],
"source": "REFERENCES_COMMIT"
},
{
"original_tag": "v0.32.0",
"range": [
{
"fixed": "5df63a7203cdb9328d824a27b682e77c7b9928be"
}
],
"source": "REFERENCES_TAG"
}
]
}