Bug 2560985 (CVE-2026-115918) - CVE-2026-115918 vllm: vLLM: Denial of service via untrusted chat template processing
Summary: CVE-2026-115918 vllm: vLLM: Denial of service via untrusted chat template pro...
Keywords:
Status: NEW
Alias: CVE-2026-115918
Product: Security Response
Classification: Other
Component: vulnerability
Version: unspecified
Hardware: All
OS: Linux
medium
medium
Target Milestone: ---
Assignee: Product Security DevOps Team
QA Contact:
URL:
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2026-10-26 17:47 UTC by OSIDB Bzimport
Modified: 2026-10-29 17:32 UTC (History)
5 users (show)

Fixed In Version:
Clone Of:
Environment:
Last Closed:
Embargoed:


Attachments (Terms of Use)

Description OSIDB Bzimport 2026-10-26 17:47:24 UTC
vLLM is an inference and serving engine for large language models. From 0.16.0 until 0.32.0, the /v1/responses endpoint allows an authenticated caller to place an attacker-controlled template in chat_template_kwargs[chat_template] when no server template is configured. ChatParams.get_apply_chat_template_kwargs() preserves that value without enforcing trust_request_chat_template, so the request can bypass the intended trust policy and cause CPU or memory exhaustion through Jinja rendering. This issue is fixed in version 0.32.0.


Note You need to log in before you can comment on or make changes to this bug.