CVE-2026-105753

The record

vLLM is an inference and serving engine for large language models. Prior to 0.28.0, the default mirrored multimodal LRU cache can commit a media hash in the frontend sender cache during multimodal rendering and before engine admission, while the engine receiver cache never receives the payload if that request is rejected. A later request reusing the same media hash causes MultiModalProcessorSenderCache to send no payload and MultiModalReceiverCache to reach an assertion with the message "Expected a cached item," producing a shared-service availability failure. This issue is fixed in version 0.28.0.

Affected packages

Related vulnerabilities

• CVE-2026-105754 vLLM is an inference and serving engine for large language models. CWE-617
• CVE-2026-98151 In the Linux kernel, the following vulnerability has been resolved: bpf: Fix REG… CWE-617
• CVE-2026-84450 libheif is a HEIF and AVIF file format decoder and encoder. CWE-617
• CVE-2026-8674 Initializing the DNS stub resolver from an /etc/resolv.conf file, or a LOCALDOMAIN… CWE-617
• CVE-2026-80274 If a BIND resolver sends a query for a DNSSEC-signed authoritative zone, and the… CWE-617
• CVE-2026-76163 If BIND is loaded with a "`named.conf`" file that contains no global "`options`" block… CWE-617
• CVE-2026-86317 A vulnerability was detected in ggml-org llama.cpp up to 0.4.0. CWE-617
• CVE-2026-20503 In Modem, there is a possible system crash due to a missing bounds check. CWE-617