CVE-2026-116178

vLLM from 0.20.0 before 0.30.0 contains a denial of service vulnerability that allows remote attackers to crash the engine by sending inconsistent multimodal metadata to the disaggregated /inference/v1/generate endpoint. Attackers can replay a valid multimodal item with repeated or overlapping...

CVE-2026-116178

A flaw was found in vLLM. This vulnerability allows a remote attacker to cause a Denial of Service DoS by crashing the inference engine. By sending crafted requests with inconsistent or overlapping multimodal metadata to the inference endpoint, an attacker can trigger an internal processing error...

CVE-2026-116175

A flaw was found in vLLM. This vulnerability allows a remote authenticated attacker to cause a Denial of Service DoS by exhausting system memory. The multimodal chat processing component fails to enforce duration limits when decompressing audio inputs. By submitting a chat completion request...

ai-captcha-bypass

AI-Powered CAPTCHA Solver This project is a Python-based command-line tool that uses large multimodal models LMMs like OpenAI's GPT-4o and Google's Gemini to automatically solve various types of CAPTCHAs. It leverages Selenium for web browser automation to interact with web pages and solve CAPTCH...

CVE-2026-116178

The vLLM library versions 0.20.0 to before 0.30.0 contains a denial of service vulnerability in the disaggregated /inference/v1/generate endpoint. The root cause is improper input validation of multimodal metadata; specifically, attackers can provide inconsistent mm_placeholders ranges or mismatc...

EUVD-2026-109439

vLLM from 0.20.0 before 0.30.0 contains a denial of service vulnerability that allows remote attackers to crash the engine by sending inconsistent multimodal metadata to the disaggregated /inference/v1/generate endpoint. Attackers can replay a valid multimodal item with repeated or overlapping...

CVE-2026-116178 vLLM 0.20.0 before 0.30.0 DoS via /inference/v1/generate mm_placeholders

vLLM from 0.20.0 before 0.30.0 contains a denial of service vulnerability that allows remote attackers to crash the engine by sending inconsistent multimodal metadata to the disaggregated /inference/v1/generate endpoint. Attackers can replay a valid multimodal item with repeated or overlapping...

Learning-to-Detect

Learning to Detect Unknown Jailbreak Attacks in Large Vision-Language Models Official implementation of “Learning to Detect Unknown Jailbreak Attacks in Large Vision-Language Models.” This repository contains the data-processing, hidden-state extraction, classifier training, safety-pattern...

bordair-multimodal

Multimodal Prompt Injection Dataset 516,588 labeled samples 251,782 attack + 251,576 benign, plus a 13,230-sample real-world validation split across five dataset versions plus external dataset ingestion, covering cross-modal, multi-turn, adversarial suffix, jailbreak template, indirect injection,...

vLLM 0.8.3 - 0.14.0 - Information Disclosure

vLLM 0.8.3 to - 0.14.1 contains an information disclosure caused by leaking a heap address in error messages from the multimodal endpoint when processing invalid images, letting remote attackers reduce ASLR entropy, exploit requires sending invalid images. id: CVE-2026-22778 info: name: vLLM 0.8....

Multimodal-Unlearnable-Examples

Multimodal Unlearnable Examples: Protecting Data against Multimodal Contrastive Learning ACM MM2024 Abstract : Multimodal contrastive learning MCL has shown remarkable advances in zero-shot classification by learning from millions of image-caption pairs crawled from the Internet. However, this...

ContextHound

ContextHound Static analysis tool that scans your codebase for LLM prompt-injection and multimodal security vulnerabilities. Runs offline, no API calls required. The ContextHound ecosystem ContextHound is available across your entire development and browsing workflow: Tool| What it does| Install...

EXPLOIT-CVE-2026-22778

CVE-2026-22778 — vLLM RCE in video processing Vulnerable lab + proof of concept for CVE-2026-22778 CVSS 9.8, an unauthenticated remote code execution chain in vLLM's multimodal ingestion path. | ---|--- CVE| CVE-2026-22778 Advisory| GHSA-4r2x-xpjr-7cvv Affected| vLLM = 0.8.3, vLLM turned...

hcaptcha-challenger

hCaptcha Challenger 🚀 Gracefully face hCaptcha challenge with multimodal large language model. Introduction Does not rely on any Tampermonkey script. Does not use any third-party anti-captcha services. Just implement some interfaces to make AI vs AI possible. Documentation: English | 简体中文 | Русск...

GHSA-GX7P-2J49-HFQ4 vLLM : DoS via unvalidated mm_processor_kwargs in multimodal image processors (Isaac, InternVL, H2OVL, etc.)

Summary API clients can supply arbitrary mmprocessorkwargs per request. vLLM forwards these values through the multimodal rendering and processing pipeline to Hugging Face and custom image/video processor constructors and calls. Resource-sensitive values such as image size, patch size, patch coun...

vLLM : DoS via unvalidated mm_processor_kwargs in multimodal image processors (Isaac, InternVL, H2OVL, etc.)

Summary API clients can supply arbitrary mmprocessorkwargs per request. vLLM forwards these values through the multimodal rendering and processing pipeline to Hugging Face and custom image/video processor constructors and calls. Resource-sensitive values such as image size, patch size, patch coun...

EUVD-2026-107006

vLLM : DoS via unvalidated mmprocessorkwargs in multimodal image processors Isaac, InternVL, H2OVL, etc...

EUVD-2026-107001

vLLM: Forged multimodal UUID + P0/P1 processor-cache eviction drift: unauthenticated remote crash of the whole vLLM engine...

GHSA-P92P-RXJ5-7P2X vLLM: Forged multimodal UUID + P0/P1 processor-cache eviction drift: unauthenticated remote crash of the whole vLLM engine

Summary Multimodal content parts imageurl, videourl, inputaudio, ... accept a user-supplied uuid field that vLLM uses verbatim as the item's cache identity mmhash. The two frontend/engine processor caches P0 "sender" cache in the API process, P1 "receiver" cache in the EngineCore process are...

vLLM: Forged multimodal UUID + P0/P1 processor-cache eviction drift: unauthenticated remote crash of the whole vLLM engine

Summary Multimodal content parts imageurl, videourl, inputaudio, ... accept a user-supplied uuid field that vLLM uses verbatim as the item's cache identity mmhash. The two frontend/engine processor caches P0 "sender" cache in the API process, P1 "receiver" cache in the EngineCore process are...