CVE-2026-58116: Inclusion of Functionality from Untrusted Control Sphere

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...

PT-2026-53876

Name of the Vulnerable Software and Affected Versions LLaMA-Factory versions prior to 0.9.6 Description Remote code execution is possible via the WebUI Chat and Training interfaces. The application fails to validate user-supplied model paths before passing them to the AutoTokenizer.from pretraine...

The vulnerability of the _process_request() function in the library for working with Large Language Models (LLM) from LLaMA-Factory allows attackers to circumvent security restrictions and execute SSRF attacks.

The vulnerability of the processrequest function in the library for working with Large Language Models LLM from LLaMA-Factory is related to the modification of requests on the server side during the processing of the /v1/chat/completions endpoint. Exploiting this vulnerability allows a malicious...

Server-side Request Forgery (SSRF)

Overview llamafactory is an Easy-to-use LLM fine-tuning framework Affected versions of this package are vulnerable to Server-side Request Forgery SSRF via the processrequest function, which processes incoming content URLs without proper validation or sanitization. An attacker can access internal...

CVE-2025-61784

LLaMA-Factory's chat API contains SSRF and LFI in the _process_request function (src/llamafactory/api/chat.py). For image_url, video_url, and audio_url, if a URL is not a base64 data URI or local file path, the code fetches the URL with requests.get(url, stream=True).raw without validation, enabl...

CVE-2025-61784 LLaMA Factory's Chat API has Critical SSRF and LFI Vulnerabilities

LLaMA-Factory is a tuning library for large language models. Prior to version 0.9.4, a Server-Side Request Forgery SSRF vulnerability in the chat API allows any authenticated user to force the server to make arbitrary HTTP requests to internal and external networks. This can lead to the exposure ...

CVE-2025-61784 LLaMA Factory's Chat API has Critical SSRF and LFI Vulnerabilities

LLaMA-Factory is a tuning library for large language models. Prior to version 0.9.4, a Server-Side Request Forgery SSRF vulnerability in the chat API allows any authenticated user to force the server to make arbitrary HTTP requests to internal and external networks. This can lead to the exposure ...

EUVD-2025-32903

LLaMA-Factory is a tuning library for large language models. Prior to version 0.9.4, a Server-Side Request Forgery SSRF vulnerability in the chat API allows any authenticated user to force the server to make arbitrary HTTP requests to internal and external networks. This can lead to the exposure ...

CVE-2025-61784: Improper Limitation of a Pathname to a Restricted Directory

LLaMA-Factory is a tuning library for large language models. Prior to version 0.9.4, a Server-Side Request Forgery SSRF vulnerability in the chat API allows any authenticated user to force the server to make arbitrary HTTP requests to internal and external networks. This can lead to the exposure ...

LLaMA-Factory 安全漏洞

LLaMA-Factory is a fine-tuned large-scale language model by a Chinese hoshi-hiyouga individual developer. A security vulnerability exists in LLaMA-Factory versions prior to 0.9.4, which stems from the processrequest function not validating or cleaning up URLs, which could lead to server-side...

EUVD-2025-12156

Malicious code in bioql PyPI...

EUVD-2025-19200

Malicious code in bioql PyPI...

EUVD-2024-3357

Malicious code in bioql PyPI...

Deserialization of Untrusted Data

Overview llamafactory is an Easy-to-use LLM fine-tuning framework Affected versions of this package are vulnerable to Deserialization of Untrusted Data via the Checkpoint path parameter in the WebUI interface during the training process. An attacker can exploit this vulnerability by supplying a...

CVE-2025-53002

LLaMA-Factory is a tuning library for large language models. A remote code execution vulnerability was discovered in LLaMA-Factory versions up to and including 0.9.3 during the LLaMA-Factory training process. This vulnerability arises because the vheadfile is loaded without proper safeguards,...

GHSA-XJ56-P8MM-QMXJ LLaMA-Factory allows Code Injection through improper vhead_file safeguards

Summary A critical remote code execution vulnerability was discovered during the Llama Factory training process. This vulnerability arises because the vheadfile is loaded without proper safeguards, allowing malicious attackers to execute arbitrary malicious code on the host system simply by passi...

LLaMA-Factory allows Code Injection through improper vhead_file safeguards

Summary A critical remote code execution vulnerability was discovered during the Llama Factory training process. This vulnerability arises because the vheadfile is loaded without proper safeguards, allowing malicious attackers to execute arbitrary malicious code on the host system simply by passi...

CVE-2025-53002

LLaMA-Factory is a tuning library for large language models. A remote code execution vulnerability was discovered in LLaMA-Factory versions up to and including 0.9.3 during the LLaMA-Factory training process. This vulnerability arises because the vheadfile is loaded without proper safeguards,...

CVE-2025-53002 LLaMA-Factory Remote Code Execution (RCE) Vulnerability

LLaMA-Factory is a tuning library for large language models. A remote code execution vulnerability was discovered in LLaMA-Factory versions up to and including 0.9.3 during the LLaMA-Factory training process. This vulnerability arises because the vheadfile is loaded without proper safeguards,...

CVE-2025-53002

Summary of CVE-2025-53002 (LLaMA-Factory) : A remote code execution vulnerability was reported in LLaMA-Factory up to version 0.9.3 during training. The root cause is loading the vhead_file without the secure parameter weights_only=True, enabling an attacker to execute arbitrary code by supplying...