CVE-2026-54236

vLLM is an inference and serving engine for large language models LLMs. Prior to 0.23.1rc0, the fix for CVE-2026-22778, which introduced a sanitizemessage helper that strips object-repr memory addresses from error messages before they reach the client, is incomplete: several response paths echo...

PT-2026-50491

Name of the Vulnerable Software and Affected Versions vLLM versions prior to 0.23.1rc0 Description An incomplete fix for a previous memory leak issue allows unauthenticated attackers to leak heap memory addresses. The system fails to properly sanitize error messages in several response paths,...