vLLM from 0.7.3 before 0.31.0 forwards the request-supplied mm_processor_kwargs.code_revision to Transformers AutoProcessor.from_pretrained, letting clients select which revision of a model's dynamic processor code is imported when trust_remote_code is enabled. An attacker controlling a revision in the served model repository can send a multimodal request to the unauthenticated /tokenize endpoint to execute arbitrary Python in the vLLM API process.
References
Configurations
No configuration.
History
21 Oct 2026, 01:17
| Type | Values Removed | Values Added |
|---|---|---|
| New CVE |
Information
Published : 2026-10-21 01:17
Updated : 2026-10-27 22:17
NVD link : CVE-2026-114053
Mitre link : CVE-2026-114053
CVE.ORG link : CVE-2026-114053
JSON object : View
Products Affected
No product.
CWE
CWE-94
Improper Control of Generation of Code ('Code Injection')
