PYSEC-2026-2018 vLLM deserialization vulnerability leading to DoS and potential RCE

Summary A memory corruption vulnerability that leading to a crash denial-of-service and potentially remote code execution RCE exists in vLLM versions 0.10.2 and later, in the Completions API endpoint. When processing user-supplied prompt embeddings, the endpoint loads serialized tensors using...

Reachable Assertion

Overview vllm is an A high-throughput and memory-efficient inference and serving engine for LLMs Affected versions of this package are vulnerable to Reachable Assertion via the /v1/completions endpoint when processing pure prompt embeds with M-RoPE models. An attacker can cause the server to cras...

CVE-2026-55514

vLLM is a library for LLM inference and serving. From 0.12.0 to before 0.24.0, sending a pure prompt embeds payload in a /v1/completions request with a model using M-RoPE causes EngineCore to fail an assertion and fatally crash, shutting down the entire server application. Any remote user who is...

PT-2026-55999

Name of the Vulnerable Software and Affected Versions vLLM versions 0.12.0 through 0.23.x Description A flaw in the EngineCore of the library occurs when a remote authorized user sends a pure prompt embeds payload to the '/v1/completions' endpoint using a model that implements M-RoPE Multimodal...

CVE-2026-56340

A flaw was found in vLLM. This vulnerability allows a remote attacker to trigger crashes or resource exhaustion, leading to a denial of service DoS. By submitting specially crafted embedding requests with malformed tensor indices, when the prompt-embeds feature is enabled, an attacker could also...

CVE-2026-48746

A flaw was found in vLLM, an inference and serving engine for large language models LLMs. This vulnerability, residing in ASGI web servers and Starlette's trust in them, allows an attacker to bypass the OpenAI API Authentication Middleware. This bypass enables unauthorized access to the API witho...

CVE-2026-41523

vLLM is an inference and serving engine for large language models LLMs. Prior to 0.22.0, an assert-based security check in vLLM's activation function loading allows any unauthenticated attacker to achieve arbitrary code execution on the server by publishing a malicious HuggingFace model, when vLL...

PYSEC-2026-2301

vLLM is an inference and serving engine for large language models LLMs. Prior to 0.22.0, vLLM's revision pinning controls do not consistently apply to all artifacts loaded for a model. A deployment that supplies --revision or --code-revision can still load dynamic code, GGUF files, image...

CVE-2026-48746

vLLM is an inference and serving engine for large language models LLMs. From 0.3.0 until 0.22.0, a vulnerability in ASGI web servers and starlette's trust on those web servers enables an authentication bypass of the OpenAI API AuthenticationMiddleware. It allows to use the API without providing t...

CVE-2026-41523

vLLM is an inference and serving engine for large language models LLMs. Prior to 0.22.0, an assert-based security check in vLLM's activation function loading allows any unauthenticated attacker to achieve arbitrary code execution on the server by publishing a malicious HuggingFace model, when vLL...

CVE-2026-54233 vLLM: OOM Denial of Service via Audio Decompression Bomb

vLLM is an inference and serving engine for large language models LLMs. Prior to 0.23.1rc0, vLLM's /v1/audio/transcriptions endpoint limits compressed upload size but not decoded PCM output. A 25MB OPUS file expands to 14.9GB of float32 PCM at decode time. This vulnerability is fixed in 0.23.1rc0...

CVE-2026-54236 vLLM: incomplete CVE-2026-22778 fix leaks PIL repr addresses via Anthropic router

vLLM is an inference and serving engine for large language models LLMs. Prior to 0.23.1rc0, the fix for CVE-2026-22778, which introduced a sanitizemessage helper that strips object-repr memory addresses from error messages before they reach the client, is incomplete: several response paths echo...

CVE-2026-48746

vLLM is an inference and serving engine for large language models LLMs. From 0.3.0 until 0.22.0, a vulnerability in ASGI web servers and starlette's trust on those web servers enables an authentication bypass of the OpenAI API AuthenticationMiddleware. It allows to use the API without providing t...

CVE-2025-71379

vLLM versions = 0.6.3 and 0.9.0 contain multiple regular expression denial of service ReDoS vulnerabilities. Several regex patterns — in vllm/lora/utils.py, the phi4mini tool parser, and the OpenAI-compatible serving chat endpoint — are susceptible to catastrophic backtracking. An attacker...

CVE-2026-56340

vLLM versions = 0.10.2 and 0.13.0 are missing sparse tensor validation in multimodal embeddings processing. Because PyTorch disables sparse tensor invariant checks by default, an attacker can submit crafted embedding requests with malformed negative or out-of-bounds tensor indices, when the...

Improper Handling of Highly Compressed Data (Data Amplification)

Overview vllm is an A high-throughput and memory-efficient inference and serving engine for LLMs Affected versions of this package are vulnerable to Improper Handling of Highly Compressed Data Data Amplification through the audio.py file. An attacker can cause excessive memory consumption by...

Incorrect Conversion between Numeric Types

Overview vllm is an A high-throughput and memory-efficient inference and serving engine for LLMs Affected versions of this package are vulnerable to Incorrect Conversion between Numeric Types in the ggmldequantize, ggmlmulmatveca8, ggmlmulmata8, and ggmlmoea8 functions when tensor dimensions are...

vLLM: temperature=NaN and temperature=Infinity bypass validation and propagate to GPU kernels

Summary All temperature validation gates use comparison operators , which silently evaluate to False for NaN and for positive Infinity in Python's IEEE 754 float semantics. Both values pass every guard and propagate to GPU sampling kernels, where they produce undefined behavior or CUDA errors tha...

HTTP Request Smuggling

Overview vllm is an A high-throughput and memory-efficient inference and serving engine for LLMs Affected versions of this package are vulnerable to HTTP Request Smuggling via improper validation of the Host header in the request scope. An attacker can gain unauthorized access to API endpoints by...

vLLM: OpenAI auth bypass

Summary A vulnerability in ASGI web servers and starlette's trust on those web servers enables an authentication bypass of the OpenAI API AuthenticationMiddleware, which was discovered during @x41sec's source code audit. It allows to use the API without providing the configured VLLMAPIKEY or...