Vulnerability · Published Jun 30, 2026 · Assigned by VulnCheck
CVE-2026-58116LLaMA-Factory 0.9.5 Remote Code Execution via WebUI Model Path
Critical 9.8
In plain English
A critical-severity code injection in LlamaFactory from hiyouga. Proof-of-concept exploit code is public. The chance it is exploited in the next 30 days is estimated at 0.9% (higher than 60% of all known vulnerabilities).
Official description
LLaMA-Factory through 0.9.5 contains a remote code execution vulnerability that allows attackers with WebUI access to execute arbitrary Python code by supplying a malicious model path in the Chat or Training interfaces. The application passes user-supplied model path input unvalidated into AutoTokenizer.from_pretrained() and AutoModel.from_pretrained() with a hardcoded trust_remote_code=True parameter, causing the Hugging Face transformers library to fetch and execute arbitrary code from a remote or local model repository with the privileges of the server process.
Affected products (as reported)
| Vendor | Product | Versions |
|---|---|---|
| hiyouga | LlamaFactory | 0 – through 0.9.5 |
Weakness: CWE-829, CWE-94 · Vector: CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H
References
- Researcher Disclosure technical-description, exploit
- https://www.vulncheck.com/advisories/llama-factory-remote-code-execution-via-webui-model-path third-party-advisory