CVE-ID

CVE-2025-46570

• CVSS Severity Rating • Fix Information • Vulnerable Software Versions • SCAP Mappings • CPE Information
Description
vLLM is an inference and serving engine for large language models (LLMs). Prior to version 0.9.0, when a new prompt is processed, if the PageAttention mechanism finds a matching prefix chunk, the prefill process speeds up, which is reflected in the TTFT (Time to First Token). These timing differences caused by matching chunks are significant enough to be recognized and exploited. This issue has been patched in version 0.9.0.
References
Note: References are provided for the convenience of the reader to help distinguish between vulnerabilities. The list is not intended to be complete.
Assigning CNA
GitHub (maintainer security advisories)
Date Record Created
20250424 Disclaimer: The record creation date may reflect when the CVE ID was allocated or reserved, and does not necessarily indicate when this vulnerability was discovered, shared with the affected vendor, publicly disclosed, or updated in CVE.
Phase (Legacy)
Assigned (20250424)
Votes (Legacy)
Comments (Legacy)
Proposed (Legacy)
N/A
This is an record on the CVE List, which provides common identifiers for publicly known cybersecurity vulnerabilities.