641 matches found
Server-side Request Forgery (SSRF)
Overview vllm is an A high-throughput and memory-efficient inference and serving engine for LLMs Affected versions of this package are vulnerable to Server-side Request Forgery SSRF via the MediaConnector class. An attacker can access internal network resources and cause system instability or...
CVE-2026-24779
CVE-2026-24779 is an SSRF vulnerability in vLLM’s MediaConnector. Before version 0.14.1, load_from_url and load_from_url_async fetch media from user-supplied URLs and validate via Python urllib urlparse, while the request is issued with requests/urllib3, whose parsing follows a different standard...
CVE-2026-24779
vLLM is an inference and serving engine for large language models LLMs. Prior to version 0.14.1, a Server-Side Request Forgery SSRF vulnerability exists in the MediaConnector class within the vLLM project's multimodal feature set. The loadfromurl and loadfromurlasync methods obtain and process...
CVE-2026-24779 vLLM vulnerable to Server-Side Request Forgery (SSRF) in `MediaConnector`
vLLM is an inference and serving engine for large language models LLMs. Prior to version 0.14.1, a Server-Side Request Forgery SSRF vulnerability exists in the MediaConnector class within the vLLM project's multimodal feature set. The loadfromurl and loadfromurlasync methods obtain and process...
PT-2026-5029
Name of the Vulnerable Software and Affected Versions vLLM versions prior to 0.14.1 Description A Server-Side Request Forgery SSRF exists in the MediaConnector class within the multimodal feature set. The load from url and load from url async functions process media from user-provided URLs using...
CVE-2026-22807 vulnerabilities
Vulnerabilities for packages: py3-vllm-cuda-12.4, tritonserver-backend-vllm-cuda-12.9, tritonserver-backend-vllm-cuda-13.0...
GHSA-2PC9-4J83-QJMR vulnerabilities
Vulnerabilities for packages: py3-vllm-cuda-12.4, tritonserver-backend-vllm-cuda-12.9, tritonserver-backend-vllm-cuda-13.0...
CVE-2026-22807
vLLM is an inference and serving engine for large language models LLMs. Starting in version 0.10.1 and prior to version 0.14.0, vLLM loads Hugging Face automap dynamic modules during model resolution without gating on trustremotecode, allowing attacker-controlled Python code in a model repo/path ...
CVE-2026-22807 vLLM affected by RCE via auto_map dynamic module loading during model initialization
vLLM is an inference and serving engine for large language models LLMs. Starting in version 0.10.1 and prior to version 0.14.0, vLLM loads Hugging Face automap dynamic modules during model resolution without gating on trustremotecode, allowing attacker-controlled Python code in a model repo/path ...
Arbitrary Code Injection
Overview vllm is an A high-throughput and memory-efficient inference and serving engine for LLMs Affected versions of this package are vulnerable to Arbitrary Code Injection via the automap process during model initialization, even when trustremotecode is false. An attacker can execute arbitrary...
GHSA-2PC9-4J83-QJMR vLLM affected by RCE via auto_map dynamic module loading during model initialization
Summary vLLM loads Hugging Face automap dynamic modules during model resolution without gating on trustremotecode, allowing attacker-controlled Python code in a model repo/path to execute at server startup. --- Impact An attacker who can influence the model repo/path local directory or remote...
vLLM affected by RCE via auto_map dynamic module loading during model initialization
Summary vLLM loads Hugging Face automap dynamic modules during model resolution without gating on trustremotecode, allowing attacker-controlled Python code in a model repo/path to execute at server startup. --- Impact An attacker who can influence the model repo/path local directory or remote...
vLLM code injection vulnerability
vLLM is an open-source LLM-based inference and service engine that features high throughput and efficient memory usage. Versions of vLLM from 0.10.1 to 0.14.0 contained a code injection vulnerability. This vulnerability stemmed from the loading of the Hugging Face automap dynamic module during...
PT-2026-3865
Name of the Vulnerable Software and Affected Versions vLLM versions 0.10.1 through 0.13.x Description vLLM is an inference and serving engine for large language models LLMs. The software loads Hugging Face auto map dynamic modules during model resolution without verifying trust remote code. This...
vLLM is vulnerable to DoS in Idefics3 vision models via image payload with ambiguous dimensions
Summary Users can crash the vLLM engine serving multimodal models that use the Idefics3 vision model implementation by sending a specially crafted 1x1 pixel image. This causes a tensor dimension mismatch that results in an unhandled runtime error, leading to complete server termination. Details T...
Allocation of Resources Without Limits or Throttling
Overview vllm is an A high-throughput and memory-efficient inference and serving engine for LLMs Affected versions of this package are vulnerable to Allocation of Resources Without Limits or Throttling in the processimageinput in the idefics3 model implementation. An attacker can cause the server...
CVE-2026-22773 vLLM is vulnerable to DoS in Idefics3 vision models via image payload with ambiguous dimensions
vLLM is an inference and serving engine for large language models LLMs. In versions from 0.6.4 to before 0.12.0, users can crash the vLLM engine serving multimodal models that use the Idefics3 vision model implementation by sending a specially crafted 1x1 pixel image. This causes a tensor dimensi...
vLLM 安全漏洞
vLLM is vLLM open source a high throughput and memory efficient inference and service engine for LLM. A security vulnerability exists in vLLM version 0.6.4 up to and including version 0.12.0, which stems from the fact that sending specially crafted 1x1 pixel images results in a tensor dimension...
GHSA-MCMC-2M55-J8JJ vLLM introduced enhanced protection for CVE-2025-62164
Summary The fix here for CVE-2025-62164 is not sufficient. The fix only disables prompt embeds by default rather than addressing the root cause, so the DoS vulnerability remains when the feature is enabled. Details vLLM's pending change attempts to fix the root cause, which is the missing sparse...