# vllm (PyPI) vulnerabilities Source: https://junglewise.ai/threats/technologies/pypi-vllm Publisher: Junglewise Threat Intelligence, https://junglewise.ai/threats Updated: 13 November 2026, 03:07 UTC. Rebuilt every hour. Data: https://junglewise.ai/threats/technologies/pypi-vllm.json License: CC BY 4.0 with a link back Junglewise Threat Intelligence has tracked 117 vulnerabilities in vllm (PyPI): 0 in the last 7 days and 51 in the last 90 days, 4 of them critical and 0 exploited in the wild. The most recent, CVE-2026-115918, was published on 26 October 2026. 20 vulnerabilities in the last four weeks, down 26% on the four before (27). ## Counts - Last 7 days: 0 - Last 30 days: 20 - Last 90 days: 51 - All time: 117 (4 critical, 0 exploited in the wild) ## About vllm (PyPI) High-throughput language model serving system. ## Latest vllm (PyPI) vulnerabilities - [CVE-2026-115918: vLLM chat template resource exhaustion DoS](https://junglewise.ai/threats/cve-2026-115918-vllm-is-an-inference-and-serving-engine-for-large-language) (medium, CVSS 6.5, published 26 Oct 2026) - [CVE-2026-115917: vLLM: Permissive default CORS origins allow cross-origin requests to the vLLM HTTP server](https://junglewise.ai/threats/cve-2026-115917-vllm-is-an-inference-and-serving-engine-for-large-language) (low, CVSS 3.4, published 26 Oct 2026) - [CVE-2026-115916: vLLM off-by-one error in tokenizer max_token_id validation](https://junglewise.ai/threats/cve-2026-115916-vllm-is-an-inference-and-serving-engine-for-large-language) (medium, CVSS 6.5, published 26 Oct 2026) - [CVE-2026-115915: vLLM image format validation bypass in media processing](https://junglewise.ai/threats/cve-2026-115915-vllm-is-an-inference-and-serving-engine-for-large-language) (high, CVSS 8.1, published 26 Oct 2026) - [CVE-2026-115914: vLLM denial of service in DeepSeek-OCR image processing](https://junglewise.ai/threats/cve-2026-115914-vllm-is-an-inference-and-serving-engine-for-large-language) (medium, CVSS 6.5, published 26 Oct 2026) - [CVE-2026-115913: vLLM Phi-4 multimodal audio encoder GPU memory exhaustion](https://junglewise.ai/threats/cve-2026-115913-vllm-is-an-inference-and-serving-engine-for-large-language) (medium, CVSS 6.5, published 26 Oct 2026) - [CVE-2026-115912: vLLM arbitrary code execution in shared-memory multimodal cache](https://junglewise.ai/threats/cve-2026-115912-vllm-is-an-inference-and-serving-engine-for-large-language) (high, CVSS 7.5, published 26 Oct 2026) - [CVE-2026-115911: vLLM chat template resource exhaustion denial of service](https://junglewise.ai/threats/cve-2026-115911-vllm-is-an-inference-and-serving-engine-for-large-language) (medium, CVSS 6.5, published 26 Oct 2026) - [CVE-2026-116178: vLLM denial of service in multimodal inference endpoint](https://junglewise.ai/threats/cve-2026-116178-vllm-from-0-20-0-before-0-30-0-contains-a-denial-of-service) (medium, CVSS 5.9, published 25 Oct 2026) - [CVE-2026-116177: vLLM uncontrolled resource consumption in Nemotron-VL processor](https://junglewise.ai/threats/cve-2026-116177-vllm-from-0-10-2-before-0-25-0-contains-an-uncontrolled-resource) (high, CVSS 7.5, published 25 Oct 2026) - [CVE-2026-116176: vLLM server-side request forgery in X-Request-Id header](https://junglewise.ai/threats/cve-2026-116176-vllm-from-0-9-2-before-0-24-0-contains-a-server-side-request) (high, CVSS 8.2, published 25 Oct 2026) - [CVE-2026-116173: vLLM Rust gRPC denial of service via output_candidates validation bypass](https://junglewise.ai/threats/cve-2026-116173-vllm-from-0-22-0-before-0-24-0-contains-a-denial-of-service) (medium, CVSS 5.9, published 25 Oct 2026) - [vLLM request-controlled mm_processor_kwargs.code_revision remote code execution](https://junglewise.ai/threats/duplicate-advisory-vllm-request-controlled-mm-processor-kwargs-code-7c748ab4) (high, CVSS 8.1, published 21 Oct 2026) - [vLLM uncontrolled resource consumption in multimodal processors](https://junglewise.ai/threats/duplicate-advisory-vllm-0-7-3-before-0-31-0-dos-via-unvalidated-mm-32363cc5) (high, CVSS 7.5, published 21 Oct 2026) - [vLLM remote denial of service through forged multimodal UUID cache desynchronization](https://junglewise.ai/threats/duplicate-advisory-vllm-vulnerable-to-remote-denial-of-service-through-c328d8d4) (high, CVSS 7.5, published 21 Oct 2026) - [CVE-2026-114055: vLLM uncontrolled resource consumption in multimodal processor](https://junglewise.ai/threats/cve-2026-114055-vllm-from-0-7-3-before-0-31-0-contains-an-uncontrolled-resource) (high, CVSS 7.5, published 21 Oct 2026) - [CVE-2026-114054: vLLM cache aliasing vulnerability in multimodal and prefix-cache](https://junglewise.ai/threats/cve-2026-114054-vllm-before-0-31-0-contains-a-cache-aliasing-vulnerability-that) (medium, CVSS 6.5, published 21 Oct 2026) - [CVE-2026-114053: vLLM remote code execution via mm_processor_kwargs.code_revision](https://junglewise.ai/threats/cve-2026-114053-vllm-from-0-7-3-before-0-31-0-forwards-the-request-supplied-mm) (high, CVSS 8.1, published 21 Oct 2026) - [CVE-2026-114052: vLLM memory exhaustion in MOSS-Audio model serving](https://junglewise.ai/threats/cve-2026-114052-vllm-after-0-24-0-and-before-0-31-0-contains-a-memory-exhaustion) (high, CVSS 7.5, published 21 Oct 2026) - [CVE-2026-114050: vLLM multimodal UUID cache poisoning remote engine crash](https://junglewise.ai/threats/cve-2026-114050-vllm-before-0-31-0-uses-a-client-supplied-multimodal-uuid) (high, CVSS 7.5, published 21 Oct 2026) - [CVE-2026-105922: vLLM denial of service in token binning penalty handler](https://junglewise.ai/threats/cve-2026-105922-a-security-flaw-has-been-discovered-in-vllm-project-vllm-up-to-0) (medium, CVSS 4.3, published 6 Oct 2026) - [CVE-2026-105760: vLLM resource exhaustion in GLMGA video backend](https://junglewise.ai/threats/cve-2026-105760-vllm-is-an-inference-and-serving-engine-for-large-language) (medium, CVSS 5.3, published 5 Oct 2026) - [CVE-2026-105759: vLLM Prometheus label injection denial of service](https://junglewise.ai/threats/cve-2026-105759-vllm-is-an-inference-and-serving-engine-for-large-language) (medium, CVSS 5.9, published 5 Oct 2026) - [CVE-2026-105758: vLLM Qwen video backend memory exhaustion via unbounded sampling parameters](https://junglewise.ai/threats/cve-2026-105758-vllm-is-an-inference-and-serving-engine-for-large-language) (medium, CVSS 5.3, published 5 Oct 2026) - [CVE-2026-105757: vLLM structured-output request handling denial of service](https://junglewise.ai/threats/cve-2026-105757-vllm-is-an-inference-and-serving-engine-for-large-language) (medium, CVSS 6.5, published 5 Oct 2026) ## Most severe vllm (PyPI) vulnerabilities - [CVE-2026-22778: vLLM remote code execution via JPEG2000 heap overflow and ASLR bypass](https://junglewise.ai/threats/cve-2026-22778-vllm-remote-code-execution-via-jpeg2000-heap-overflow-and-aslr) (critical, CVSS 9.8, published 2 Feb 2026) - [CVE-2024-9053: vLLM remote code execution via pickle deserialization in AsyncEngineRPCServer](https://junglewise.ai/threats/cve-2024-9053-vllm-remote-code-execution-via-pickle-deserialization-in) (critical, CVSS 9.8, published 20 Mar 2025) - [CVE-2025-47277: vLLM remote code execution via unsafe deserialization in PyNcclPipe](https://junglewise.ai/threats/cve-2025-47277-vllm-remote-code-execution-via-unsafe-deserialization-in) (critical, CVSS 9.8, published 20 May 2025) - [CVE-2026-48746: vLLM authentication bypass in OpenAI API via Host header injection](https://junglewise.ai/threats/cve-2026-48746-vllm-authentication-bypass-in-openai-api-via-host-header) (critical, CVSS 9.1, published 22 Jun 2026) - [CVE-2026-27893: vLLM remote code execution via hardcoded remote code trust in models](https://junglewise.ai/threats/cve-2026-27893-vllm-remote-code-execution-via-hardcoded-remote-code-trust-in) (high, CVSS 8.8, published 27 Mar 2026) - [CVE-2025-62164: vLLM unsafe deserialization in Completions API prompt embeddings](https://junglewise.ai/threats/cve-2025-62164-vllm-unsafe-deserialization-in-completions-api-prompt-embeddings) (high, CVSS 8.8, published 21 Nov 2025) - [CVE-2026-22807: vLLM arbitrary code execution via auto_map dynamic module loading](https://junglewise.ai/threats/cve-2026-22807-vllm-arbitrary-code-execution-via-auto-map-dynamic-module-loading) (high, CVSS 8.8, published 21 Jan 2026) - [CVE-2026-56340: vLLM improper input validation in multimodal embeddings](https://junglewise.ai/threats/cve-2026-56340-vllm-improper-input-validation-in-multimodal-embeddings) (high, CVSS 8.8, published 20 Jun 2026) - [CVE-2026-54232: vLLM dependency confusion in Dockerfile via flashinfer-jit-cache](https://junglewise.ai/threats/cve-2026-54232-vllm-dependency-confusion-in-dockerfile-via-flashinfer-jit-cache) (high, CVSS 8.8, published 22 Jun 2026) - [CVE-2026-116176: vLLM server-side request forgery in X-Request-Id header](https://junglewise.ai/threats/cve-2026-116176-vllm-from-0-9-2-before-0-24-0-contains-a-server-side-request) (high, CVSS 8.2, published 25 Oct 2026) ## Vulnerabilities per week 20 vulnerabilities in the last four weeks, down 26% on the four before (27). | Week of | Vulns | Critical | Exploited | | --- | --- | --- | --- | | 17 Aug 2026 | 2 | 0 | 0 | | 24 Aug 2026 | 1 | 0 | 0 | | 31 Aug 2026 | 0 | 0 | 0 | | 7 Sep 2026 | 1 | 0 | 0 | | 14 Sep 2026 | 7 | 0 | 0 | | 21 Sep 2026 | 9 | 0 | 0 | | 28 Sep 2026 | 1 | 0 | 0 | | 5 Oct 2026 | 10 | 0 | 0 | | 12 Oct 2026 | 0 | 0 | 0 | | 19 Oct 2026 | 12 | 0 | 0 | | 26 Oct 2026 | 8 | 0 | 0 | | 2 Nov 2026 | 0 | 0 | 0 | | 9 Nov 2026 | 0 | 0 | 0 | ## Other PyPI technologies - [tensorflow (PyPI)](https://junglewise.ai/threats/technologies/pypi-tensorflow): 435 vulnerabilities - [tensorflow-cpu (PyPI)](https://junglewise.ai/threats/technologies/tensorflow-cpu): 430 vulnerabilities - [tensorflow-gpu (PyPI)](https://junglewise.ai/threats/technologies/tensorflow-gpu): 427 vulnerabilities - [open-webui (PyPI)](https://junglewise.ai/threats/technologies/open-webui): 177 vulnerabilities - [Django (PyPI)](https://junglewise.ai/threats/technologies/django): 173 vulnerabilities - [apache-airflow (PyPI)](https://junglewise.ai/threats/technologies/apache-airflow): 159 vulnerabilities - [pillow (PyPI)](https://junglewise.ai/threats/technologies/pillow): 125 vulnerabilities - [praisonai (PyPI)](https://junglewise.ai/threats/technologies/pypi-praisonai): 120 vulnerabilities - [plone (PyPI)](https://junglewise.ai/threats/technologies/pypi-plone): 101 vulnerabilities - [mlflow (PyPI)](https://junglewise.ai/threats/technologies/mlflow): 88 vulnerabilities - [picklescan (PyPI)](https://junglewise.ai/threats/technologies/picklescan): 86 vulnerabilities - [exiv2 (PyPI)](https://junglewise.ai/threats/technologies/exiv2): 85 vulnerabilities ## Sources NVD, GitHub Security Advisories, OSV, the CISA Known Exploited Vulnerabilities catalog, FIRST EPSS and vendor advisories. Rebuilt every hour. Cite as: Junglewise Threat Intelligence, "vllm (PyPI) vulnerabilities", https://junglewise.ai/threats/technologies/pypi-vllm, 13 November 2026.