vLLM from 0.10.1 before 0.31.0 forwards a request-controlled mm_processor_kwargs.max_patches value to the Hugging Face processor for Cohere2 Vision models without a server-side upper bound, causing quadratic CPU work during multimodal preprocessing. Remote clients can send small image chat-completion requests with large, varying max_patches values to bypass the LRU cache, exhaust preprocessing workers, and deny service to other clients.
{
"cna_assigner": "VulnCheck",
"cwe_ids": [
"CWE-400"
],
"osv_generated_from": "https://github.com/CVEProject/cvelistV5/tree/main/cves/2026/126xxx/CVE-2026-126397.json"
}{
"extracted_events": [
{
"range": [
{
"introduced": "0.10.1"
},
{
"fixed": "0.31.0"
}
],
"source": "AFFECTED_FIELD",
"strategy": "StandardRange"
},
{
"range": [
{
"fixed": "f6326f53bda46898a331c2d24500332c285d9a2b"
}
],
"source": "REFERENCES_COMMIT"
}
]
}