CVE-2025-62426: vLLM vulnerable to DoS via large Chat Completion or Tokenization requests with specially crafted `chat_template_kwargs`
(updated )
The /v1/chat/completions and /tokenize endpoints allow a chat_template_kwargs request parameter that is used in the code before it is properly validated against the chat template. With the right chat_template_kwargs parameters, it is possible to block processing of the API server for long periods of time, delaying all other requests
References
- github.com/advisories/GHSA-69j4-grxj-j64p
- github.com/vllm-project/vllm
- github.com/vllm-project/vllm/blob/2a6dc67eb520ddb9c4138d8b35ed6fe6226997fb/vllm/entrypoints/chat_utils.py
- github.com/vllm-project/vllm/blob/2a6dc67eb520ddb9c4138d8b35ed6fe6226997fb/vllm/entrypoints/openai/serving_engine.py
- github.com/vllm-project/vllm/commit/3ada34f9cb4d1af763fdfa3b481862a93eb6bd2b
- github.com/vllm-project/vllm/pull/27205
- github.com/vllm-project/vllm/security/advisories/GHSA-69j4-grxj-j64p
- nvd.nist.gov/vuln/detail/CVE-2025-62426
Code Behaviors & Features
Detect and mitigate CVE-2025-62426 with GitLab Dependency Scanning
Secure your software supply chain by verifying that all open source dependencies used in your projects contain no disclosed vulnerabilities. Learn more about Dependency Scanning →