CVE-2024-12055 DoS using malicious gguf model file in ollama/ollama

A vulnerability in Ollama versions =0.3.14 allows a malicious user to create a customized gguf model file that can be uploaded to the public Ollama server. When the server processes this malicious model, it crashes, leading to a Denial of Service DoS attack. The root cause of the issue is an...

CVE-2024-12029 Remote Code Execution via Model Deserialization in invoke-ai/invokeai

A remote code execution vulnerability exists in invoke-ai/invokeai versions 5.3.1 through 5.4.2 via the /api/v2/models/install API. The vulnerability arises from unsafe deserialization of model files using torch.load without proper validation. Attackers can exploit this by embedding malicious cod...

Timing Side-channel Attacks

postquantumfeldmanvss is vulnerable to Timing side-channel attacks. The vulnerability is due to Python's non-constant-time execution model, which causes execution time variations in the findsecurepivot and securematrixsolve functions, allowing attackers to infer secret information through precise...

dify 代码问题漏洞

dify is an open source LLM application development platform from LangGenius Open Source. A code issue vulnerability exists in version 0.10.1 of dify, which stems from an unvalidated URL and could lead to a server-side request forgery attack...

Open WebUI 安全漏洞

Open WebUI is an extensible, feature-rich, user-friendly self-hosted WebUI from Open WebUI open source. A security vulnerability exists in Open WebUI version 0.3.8, which stems from an improperly handled file path in the downloadmodel endpoint, which could lead to arbitrary file writes...

编号撤回

Ollama is a large language model that can be started and run locally by Ollama Open Source. This CVE number has been withdrawn...

Invoke 安全漏洞

Invoke is a leading creative engine for stabilizing diffusion models open-sourced by InvokeAI. A security vulnerability exists in Invoke versions 5.3.1 through 5.4.2, which stems from improper deserialization of model files and could lead to remote code execution...

PT-2025-12312 · Ollama · Ollama

Name of the Vulnerable Software and Affected Versions: ollama/ollama versions prior to 0.3.14 Description: A malicious user can upload and create a customized GGUF model file on the Ollama server, leading to a division by zero error in the ggufPadding function. This error causes the server to...

H2O 安全漏洞

H2O is an in-memory platform for distributed, scalable machine learning open-sourced by H2O.ai. A security vulnerability exists in H2O version 3.46.0 that stems from a model export endpoint that does not restrict the export location, which could lead to arbitrary file overwrites...

ComfyUI 跨站请求伪造漏洞

ComfyUI is one of the most powerful and modular diffusion model GUIs and backends from comfyanonymous individual developers. A cross-site request forgery vulnerability exists in ComfyUI v0.2.2 and prior versions, which stems from insufficient CSRF protection and could lead to unauthorized API...

PT-2025-12118 · Ollama · Ollama

Name of the Vulnerable Software and Affected Versions: Ollama versions =0.3.14 Description: A malicious user can create a customized gguf model file that can be uploaded to the public Ollama server, causing it to crash and leading to a Denial of Service DoS attack. The root cause of the issue is ...

PT-2025-12034 · Unknown · Lunary-Ai/Lunary

Name of the Vulnerable Software and Affected Versions: lunary-ai/lunary version 1.5.0 Description: The issue is related to improper privilege management in the models.ts file, which allows users with viewer roles to modify models owned by others. The "PATCH" endpoint for models lacks appropriate...

Ollama 安全漏洞

Ollama is an Ollama open source large-scale language model that can be started and run locally. A security vulnerability exists in Ollama version 0.3.14 and earlier, which stems from the fact that uploading a custom GGUF model file may cause the server to allocate unlimited memory, leading to a...

Lunary 安全漏洞

lunary is lunary open source a production toolkit for LLM . An information disclosure vulnerability exists in lunary that stems from a GET /projects API endpoint exposing all project public and private API keys to users with least privileges, which can be exploited by an attacker to obtain...

Lunary 安全漏洞

Lunary is Lunary open source a production toolkit for LLM . Lunary afc5df4 version of a security vulnerability , the vulnerability stems from a flaw in the permission checking mechanism , an attacker can use this vulnerability to cause unauthorized access to sensitive endpoints...

H2O Vulnerable to Arbitrary File Overwrite via File Export

In h2oai/h2o-3 version 3.46.0, the endpoint for exporting models does not restrict the export location, allowing an attacker to export a model to any file in the server's file structure, thereby overwriting it. This vulnerability can be exploited to overwrite any file on the target server with a...

PT-2025-12171

Name of the Vulnerable Software and Affected Versions h2oai/h2o-3 version 3.46.0 Description The endpoint for exporting models does not restrict the export location, allowing an attacker to export a model to any file in the server's file structure, thereby overwriting it. This issue can be...

PT-2025-12173

Name of the Vulnerable Software and Affected Versions open-webui/open-webui version 0.3.8 Description An arbitrary file write vulnerability exists in the download model endpoint. When deployed on Windows, the application improperly handles file paths, allowing an attacker to manipulate the file...

编号撤回

vLLM is vLLM open source a high throughput and memory efficient inference and service engine for LLM. This CVE number has been withdrawn...

Ollama 数字错误漏洞

Ollama is an Ollama open source large-scale language model that can be started and run locally. A numeric error vulnerability exists in Ollama version v0.3.3, which stems from a blockcount type specialization when importing a GGUF model that could result in a divide-by-zero error, triggering a...