Lucene search
+L

747 matches found

Chainguard
Chainguard
added 2025/05/30 1:15 a.m.14 views

GHSA-VRQ3-R879-7M65 vulnerabilities

Vulnerabilities for packages: tritonserver-backend-vllm, py3.10-vllm-cuda-11.8...

5.3AI score
SaveExploits0
Chainguard
Chainguard
added 2025/05/30 1:15 a.m.11 views

GHSA-W6Q7-J642-7C25 vulnerabilities

Vulnerabilities for packages: tritonserver-backend-vllm, py3.10-vllm-cuda-11.8...

5.3AI score
SaveExploits0
Chainguard
Chainguard
added 2025/05/30 1:15 a.m.43 views

CVE-2025-46570 vulnerabilities

Vulnerabilities for packages: tritonserver-backend-vllm, py3.10-vllm-cuda-11.8...

2.6CVSS4.6AI score0.00293EPSS
SaveExploits0
CNNVD
CNNVD
added 2025/05/30 12:00 a.m.13 views

vLLM 安全漏洞

vLLM is a high throughput and memory efficient inference and service engine for LLM from the vLLM open source. A security vulnerability exists in vLLM versions prior to 0.8.0 through 0.9.0, which stems from the fact that supplying an invalid regular expression when using structured output may...

6.5CVSS6.3AI score0.00475EPSS
SaveExploits0References5
OSV
OSV
added 2025/05/29 5:15 p.m.10 views

PYSEC-2025-43

vLLM is an inference and serving engine for large language models LLMs. In versions starting from 0.7.0 to before 0.9.0, in the file vllm/multimodal/hasher.py, the MultiModalHasher class has a security and data integrity issue in its image hashing method. Currently, it serializes PIL.Image.Image...

7.3CVSS7.2AI score0.00316EPSS
SaveExploits0References3
NVD
NVD
added 2025/05/29 5:15 p.m.18 views

CVE-2025-46570

vLLM is an inference and serving engine for large language models LLMs. Prior to version 0.9.0, when a new prompt is processed, if the PageAttention mechanism finds a matching prefix chunk, the prefill process speeds up, which is reflected in the TTFT Time to First Token. These timing differences...

2.6CVSS0.00293EPSS
SaveExploits0References3
NVD
NVD
added 2025/05/29 5:15 p.m.23 views

CVE-2025-46722

vLLM is an inference and serving engine for large language models LLMs. In versions starting from 0.7.0 to before 0.9.0, in the file vllm/multimodal/hasher.py, the MultiModalHasher class has a security and data integrity issue in its image hashing method. Currently, it serializes PIL.Image.Image...

7.3CVSS0.00316EPSS
SaveExploits0References3
Snyk
Snyk
added 2025/05/29 4:43 p.m.10 views

Timing Attack

Overview vllm is an A high-throughput and memory-efficient inference and serving engine for LLMs Affected versions of this package are vulnerable to Timing Attack due to the PageAttention mechanism. An attacker can observe timing differences to infer details about the processed data by analyzing...

6.3CVSS6.9AI score0.00293EPSS
SaveExploits0References2
CVE
CVE
added 2025/05/29 4:36 p.m.205 views

CVE-2025-46722

The CVE-2025-46722 entry concerns vLLM (versions 0.7.0–0.8.x) where MultiModalHasher in vllm/multimodal/hasher.py hashes PIL.Image.Image objects using only obj.tobytes(). This excludes image metadata (width, height, mode), enabling two images with identical pixel data but different shapes to yiel...

7.3CVSS4.6AI score0.00316EPSS
SaveExploits0References3Affected Software1
Vulnrichment
Vulnrichment
added 2025/05/29 4:36 p.m.14 views

CVE-2025-46722 vLLM has a Weakness in MultiModalHasher Image Hashing Implementation

vLLM is an inference and serving engine for large language models LLMs. In versions starting from 0.7.0 to before 0.9.0, in the file vllm/multimodal/hasher.py, the MultiModalHasher class has a security and data integrity issue in its image hashing method. Currently, it serializes PIL.Image.Image...

4.2CVSS4.6AI score0.00316EPSS
SaveExploits0References3
OSV
OSV
added 2025/05/29 4:36 p.m.11 views

CVE-2025-46722 vLLM has a Weakness in MultiModalHasher Image Hashing Implementation

vLLM is an inference and serving engine for large language models LLMs. In versions starting from 0.7.0 to before 0.9.0, in the file vllm/multimodal/hasher.py, the MultiModalHasher class has a security and data integrity issue in its image hashing method. Currently, it serializes PIL.Image.Image...

4.2CVSS6.4AI score
SaveExploits0References5
Cvelist
Cvelist
added 2025/05/29 4:36 p.m.32 views

CVE-2025-46722 vLLM has a Weakness in MultiModalHasher Image Hashing Implementation

vLLM is an inference and serving engine for large language models LLMs. In versions starting from 0.7.0 to before 0.9.0, in the file vllm/multimodal/hasher.py, the MultiModalHasher class has a security and data integrity issue in its image hashing method. Currently, it serializes PIL.Image.Image...

4.2CVSS0.00316EPSS
SaveExploits0References3
CVE
CVE
added 2025/05/29 4:32 p.m.210 views

CVE-2025-46570

The CVE-2025-46570 entry concerns vLLM (inference/serving engine). The concrete detail across connected records shows a vulnerability in the PageAttention-based prefill path: when a new prompt is processed, a matching prefix chunk can accelerate prefill, creating timing differences (TTFT) that co...

2.6CVSS3.6AI score0.00293EPSS
SaveExploits0References3Affected Software1
Vulnrichment
Vulnrichment
added 2025/05/29 4:32 p.m.21 views

CVE-2025-46570 vLLM’s Chunk-Based Prefix Caching Vulnerable to Potential Timing Side-Channel

vLLM is an inference and serving engine for large language models LLMs. Prior to version 0.9.0, when a new prompt is processed, if the PageAttention mechanism finds a matching prefix chunk, the prefill process speeds up, which is reflected in the TTFT Time to First Token. These timing differences...

2.6CVSS3.8AI score0.00293EPSS
SaveExploits0References3
Cvelist
Cvelist
added 2025/05/29 4:32 p.m.36 views

CVE-2025-46570 vLLM’s Chunk-Based Prefix Caching Vulnerable to Potential Timing Side-Channel

vLLM is an inference and serving engine for large language models LLMs. Prior to version 0.9.0, when a new prompt is processed, if the PageAttention mechanism finds a matching prefix chunk, the prefill process speeds up, which is reflected in the TTFT Time to First Token. These timing differences...

2.6CVSS0.00293EPSS
SaveExploits0References3
Chainguard
Chainguard
added 2025/05/29 1:15 a.m.12 views

GHSA-HJQ4-87XH-G4FV vulnerabilities

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

7.5AI score
SaveExploits0
CNNVD
CNNVD
added 2025/05/29 12:00 a.m.9 views

vLLM 安全漏洞

vLLM is a high throughput and memory efficient inference and service engine for LLM from the vLLM open source. A security vulnerability exists in vLLM versions prior to 0.7.0 through 0.9.0, which stems from the use of only raw pixel data without including metadata in the image hash method, which...

7.3CVSS6.3AI score0.00316EPSS
SaveExploits0References4
Github Security Blog
Github Security Blog
added 2025/05/28 7:42 p.m.29 views

vLLM Tool Schema allows DoS via Malformed pattern and type Fields

Summary The vLLM backend used with the /v1/chat/completions OpenAPI endpoint fails to validate unexpected or malformed input in the "pattern" and "type" fields when the tools functionality is invoked. These inputs are not validated before being compiled or parsed, causing a crash of the inference...

6.5CVSS6.6AI score0.00517EPSS
SaveExploits1References7Affected Software1
OSV
OSV
added 2025/05/28 7:42 p.m.11 views

GHSA-VRQ3-R879-7M65 vLLM Tool Schema allows DoS via Malformed pattern and type Fields

Summary The vLLM backend used with the /v1/chat/completions OpenAPI endpoint fails to validate unexpected or malformed input in the "pattern" and "type" fields when the tools functionality is invoked. These inputs are not validated before being compiled or parsed, causing a crash of the inference...

6.5CVSS7AI score0.00517EPSS
SaveExploits1References7
Snyk
Snyk
added 2025/05/28 7:42 p.m.10 views

Uncaught Exception

Overview vllm is an A high-throughput and memory-efficient inference and serving engine for LLMs Affected versions of this package are vulnerable to Uncaught Exception through the guidedregex parameter when using xgrammar validation. An attacker can cause the application to crash by sending an...

8.7CVSS7AI score0.00475EPSS
SaveExploits0References2
Rows per page
Query Builder