CVE-2025-62426 vulnerabilities

Vulnerabilities for packages: tritonserver-backend-vllm-cuda-12.9...

CVE-2025-62372 vulnerabilities

Vulnerabilities for packages: tritonserver-backend-vllm-cuda-12.9...

GHSA-PMQF-X6X8-P7QW vulnerabilities

Vulnerabilities for packages: tritonserver-backend-vllm-cuda-12.9...

CVE-2025-66448: Improper Control of Generation of Code

vLLM is an inference and serving engine for large language models LLMs. Prior to 0.11.1, vllm has a critical remote code execution vector in a config class named NemotronNanoVLConfig. When vllm loads a model config that contains an automap entry, the config class resolves that mapping with...

The vulnerability of the vLLM library for working with large language models involves unverified array indexing, which allows attackers to trigger service interruptions.

The vulnerability of the vLLM library for working with large language models is related to unverified array indexing. Exploiting this vulnerability could allow a malicious actor to cause service failures...

The vulnerability of the Completions API component in the LLM library vLLM allows a attacker to trigger a service failure and execute arbitrary code.

The vulnerability of the Completions API component in the vLLM language model processing library is related to writing beyond the buffer boundaries in memory. Exploiting this vulnerability could allow an attacker to trigger a service failure and execute arbitrary code...

The vulnerability of the vLLM library for working with large language models, related to unlimited resource distribution, allows a hacker to trigger a service failure.

The vulnerability of the vLLM library for working with large language models is related to the unlimited distribution of resources during the processing of the chattemplatekwargs parameter. Exploiting this vulnerability could allow a malicious actor to cause service failures...

Server-Side Request Forgery (SSRF)

vllm is vulnerable to Server-Side Request Forgery SSRF. The vulnerability is due to insufficient restrictions on user-supplied URLs in the MediaConnector class’s loadfromurl and loadfromurlasync methods, which allows an attacker to coerce the server into making arbitrary internal network requests...

CVE-2025-62164

vLLM is an inference and serving engine for large language models LLMs. From versions 0.10.2 to before 0.11.1, a memory corruption vulnerability could lead to a crash denial-of-service and potentially remote code execution RCE, exists in the Completions API endpoint. When processing user-supplied...

CVE-2025-62372

vLLM is an inference and serving engine for large language models LLMs. From version 0.5.5 to before 0.11.1, users can crash the vLLM engine serving multimodal models by passing multimodal embedding inputs with correct ndim but incorrect shape e.g. hidden dimension is wrong, regardless of whether...

CVE-2025-62372 vLLM vulnerable to DoS with incorrect shape of multimodal embedding inputs

vLLM is an inference and serving engine for large language models LLMs. From version 0.5.5 to before 0.11.1, users can crash the vLLM engine serving multimodal models by passing multimodal embedding inputs with correct ndim but incorrect shape e.g. hidden dimension is wrong, regardless of whether...

CVE-2025-62372 vLLM vulnerable to DoS with incorrect shape of multimodal embedding inputs

vLLM is an inference and serving engine for large language models LLMs. From version 0.5.5 to before 0.11.1, users can crash the vLLM engine serving multimodal models by passing multimodal embedding inputs with correct ndim but incorrect shape e.g. hidden dimension is wrong, regardless of whether...

CVE-2025-62426 vLLM vulnerable to DoS via large Chat Completion or Tokenization requests with specially crafted `chat_template_kwargs`

vLLM is an inference and serving engine for large language models LLMs. From version 0.5.5 to before 0.11.1, the /v1/chat/completions and /tokenize endpoints allow a chattemplatekwargs request parameter that is used in the code before it is properly validated against the chat template. With the...

CVE-2025-62426 vLLM vulnerable to DoS via large Chat Completion or Tokenization requests with specially crafted `chat_template_kwargs`

vLLM is an inference and serving engine for large language models LLMs. From version 0.5.5 to before 0.11.1, the /v1/chat/completions and /tokenize endpoints allow a chattemplatekwargs request parameter that is used in the code before it is properly validated against the chat template. With the...

CVE-2025-62426 vLLM vulnerable to DoS via large Chat Completion or Tokenization requests with specially crafted `chat_template_kwargs`

vLLM is an inference and serving engine for large language models LLMs. From version 0.5.5 to before 0.11.1, the /v1/chat/completions and /tokenize endpoints allow a chattemplatekwargs request parameter that is used in the code before it is properly validated against the chat template. With the...

CVE-2025-62164

The CVE affects vLLM (inference/serving engine) before 0.11.1, where the Completions API loads user-supplied prompt embeddings with torch.load() lacking proper validation. A PyTorch 2.8.0 change disables sparse-tensor invariants checks, allowing crafted tensors to bypass bounds checks and trigger...

CVE-2025-62164 VLLM deserialization vulnerability leading to DoS and potential RCE

vLLM is an inference and serving engine for large language models LLMs. From versions 0.10.2 to before 0.11.1, a memory corruption vulnerability could lead to a crash denial-of-service and potentially remote code execution RCE, exists in the Completions API endpoint. When processing user-supplied...

CVE-2025-62164 VLLM deserialization vulnerability leading to DoS and potential RCE

vLLM is an inference and serving engine for large language models LLMs. From versions 0.10.2 to before 0.11.1, a memory corruption vulnerability could lead to a crash denial-of-service and potentially remote code execution RCE, exists in the Completions API endpoint. When processing user-supplied...

CVE-2025-62164 VLLM deserialization vulnerability leading to DoS and potential RCE

vLLM is an inference and serving engine for large language models LLMs. From versions 0.10.2 to before 0.11.1, a memory corruption vulnerability could lead to a crash denial-of-service and potentially remote code execution RCE, exists in the Completions API endpoint. When processing user-supplied...

CVE-2025-62372: Improper Validation of Array Index

vLLM is an inference and serving engine for large language models LLMs. From version 0.5.5 to before 0.11.1, users can crash the vLLM engine serving multimodal models by passing multimodal embedding inputs with correct ndim but incorrect shape e.g. hidden dimension is wrong, regardless of whether...