vllm-project/vllm version 0.14.1 contains a vulnerability where the `trust_remote_code=True` parameter is hardcoded in two model implementation files (`vllm/model_executor/models/nemotron_vl.py` and `vllm/model_executor/models/kimi_k25.py`). This bypasses the user's explicit `--trust-remote-code=False` setting, enabling remote code execution via malicious HuggingFace model repositories. This issue is an incomplete fix for CVE-2025-66448 and CVE-2026-22807, as it affects separate code paths in model implementation files. Deployments loading NemotronVL or KimiK25 models are particularly impacted.
References
Configurations
No configuration.
History
28 May 2026, 21:16
| Type | Values Removed | Values Added |
|---|---|---|
| References | () https://huntr.com/bounties/97f706f7-a852-49b2-a4eb-76811e611daf - |
28 May 2026, 19:16
| Type | Values Removed | Values Added |
|---|---|---|
| New CVE |
Information
Published : 2026-05-28 19:16
Updated : 2026-05-29 15:39
NVD link : CVE-2026-4944
Mitre link : CVE-2026-4944
CVE.ORG link : CVE-2026-4944
JSON object : View
Products Affected
No product.
CWE
CWE-22
Improper Limitation of a Pathname to a Restricted Directory ('Path Traversal')
