Skip to main content

Vllm

45 CVEs product

Monthly

CVE-2026-55514 Jul 06, 20:07 PyPI HIGH PATCH GHSA This Week

Denial of service in vLLM 0.12.0 through 0.23.x lets any authorized API caller crash the entire inference server by submitting a pure prompt-embeddings payload to the /v1/completions endpoint when a model using M-RoPE (multimodal rotary position embedding) is loaded. The malformed request trips a reachable assertion in the EngineCore process, which terminates the whole server rather than rejecting the single request. There is no public exploit identified at time of analysis and it is not listed in CISA KEV; the flaw is fixed in vLLM 0.24.0.

Denial Of Service Red Hat Vllm Vllm Project
NVD GitHub
CVSS 4.0
7.1
EPSS
0.3%
CVE-2026-55574 Jul 06, 20:05 PyPI HIGH PATCH GHSA This Week

Denial of service in vLLM inference servers prior to 0.24.0 allows remote unauthenticated attackers to hang an inference worker indefinitely by submitting a single request with an adversarial regular expression via the structured_outputs.regex API parameter. The pattern is passed to grammar compiler backends (xgrammar with no guard, outlines with structural-but-not-complexity validation) where nested quantifiers trigger exponential state-space expansion (ReDoS). No public exploit identified at time of analysis, though the trivial request-based trigger makes weaponization straightforward.

Denial Of Service Red Hat Vllm Vllm Project
NVD GitHub
CVSS 4.0
8.7
EPSS
0.3%
CVE-2026-54234 Jul 06, 19:49 PyPI HIGH POC PATCH GHSA This Week

Denial of service in the vLLM LLM inference server (all versions prior to 0.24.0) allows a remote client to crash the shared engine worker by sending a specific multi-request speculative decoding workload. The rejection sampler produces a recovered token equal to the vocabulary-size boundary, which is coerced to -1, written back into the drafter's input ids, and later dereferenced by the embedding/attention path, triggering a GPU device-side assertion that kills the worker. There is no public exploit identified at time of analysis and this CVE is not in CISA KEV; per CVSS the impact is availability-only (C:N/I:N/A:H) with a low EPSS profile expected for a crash-only bug.

Denial Of Service Red Hat Vllm Vllm Project
NVD GitHub
CVSS 3.1
7.5
EPSS
0.3%
CVE-2026-54232 Jun 22, 22:16 PyPI HIGH PATCH This Week

Remote code execution in vLLM versions prior to 0.22.1 allows attackers to backdoor production LLM inference deployments through a dependency confusion attack in the project's Dockerfile. Because flashinfer-jit-cache was pulled via --extra-index-url with UV_INDEX_STRATEGY=unsafe-best-match while the name remained unregistered on PyPI, any attacker who claimed the name on PyPI with a higher version would have their code executed as root during every Docker build. No public exploit identified at time of analysis, but the supply-chain primitive is well understood and trivially weaponizable.

RCE Docker Red Hat Vllm Vllm Project
NVD GitHub VulDB
CVSS 3.1
8.8
EPSS
0.3%
CVE-2026-56340 Jun 20, 18:27 PyPI HIGH PATCH This Week

Denial of service and potential memory corruption in vLLM versions 0.10.2 through 0.12.x stems from missing sparse tensor validation in multimodal embeddings processing, allowing authenticated remote users to submit crafted prompt-embedding requests with malformed tensor indices. Because PyTorch disables sparse tensor invariant checks by default, attackers can crash the inference server or exhaust resources, with potential out-of-bounds or write-what-where memory corruption. No public exploit identified at time of analysis; this continuation of CVE-2025-62164 addresses the root cause that the prior fix only masked by disabling the feature by default.

Buffer Overflow Denial Of Service Red Hat Vllm
NVD GitHub VulDB
CVSS 4.0
8.7
EPSS
0.3%
CVE-2026-34760 Apr 02, 18:59 PyPI HIGH PATCH GHSA This Week

vLLM versions 0.5.5 through 0.17.x use incorrect mono audio downmixing via numpy.mean instead of the ITU-R BS.775-4 weighted standard, causing audio processed by AI models to diverge from human perception. An authenticated remote attacker with low privileges can exploit this inconsistency to manipulate audio-based model outputs or infer mismatches between expected and actual audio processing, affecting integrity of audio-driven inference pipelines. The vulnerability has been patched in vLLM 0.18.0.

Information Disclosure Vllm Vllm Project
NVD GitHub VulDB
CVSS 3.1
7.1
EPSS
0.1%
CVE-2026-27893 Mar 27, 00:16 PyPI HIGH PATCH GHSA This Week

Remote code execution is possible in vLLM inference and serving engine versions 0.10.1 through 0.17.x due to hardcoded trust_remote_code=True settings in two model implementation files that override users' explicit --trust-remote-code=False security configuration. Attackers can exploit this by hosting malicious model repositories that execute arbitrary code when loaded by vLLM, even when users have intentionally disabled remote code trust for security. Version 0.18.0 patches this vulnerability, with no public exploit identified at time of analysis and a CVSS score of 8.8 requiring user interaction to trigger.

RCE Vllm
NVD GitHub VulDB
CVSS 3.1
8.8
EPSS
0.0%
CVE-2026-25960 Mar 09, 21:16 PyPI CRITICAL PATCH Act Now

Server-Side Request Forgery in vLLM's multimodal MediaConnector allows remote attackers to coerce the inference server into fetching attacker-chosen internal URLs, bypassing the allowed_media_domains allowlist that was added to fix CVE-2026-24779. The bypass exploits a backslash-before-@ parsing disagreement between urllib3.util.parse_url (validation layer) and aiohttp/yarl (HTTP client), so a URL the allowlist reads as a permitted host is actually fetched from an internal target. It carries a CVSS 9.8 rating, a vendor patch and Red Hat advisory are available, and a working bypass payload is published in the GHSA advisory and fix test suite, though EPSS is very low (0.02%, 3rd percentile) and it is not on CISA KEV.

SSRF Vllm
NVD GitHub VulDB
CVSS 3.1
9.8
EPSS
0.0%
CVE-2026-22778 Feb 02, 23:16 PyPI CRITICAL POC PATCH GHSA Act Now

Information exposure in vLLM inference engine versions 0.8.3 to before 0.14.1. Invalid image requests to the multimodal endpoint cause sensitive data logging. Patch available.

RCE Vllm
NVD GitHub VulDB
CVSS 3.1
9.8
EPSS
0.1%
CVE-2026-24779 Jan 27, 22:15 PyPI HIGH POC PATCH GHSA This Week

vLLM before version 0.14.1 contains a server-side request forgery vulnerability in the MediaConnector class where inconsistent URL parsing between libraries allows attackers to bypass host restrictions and force the server to make arbitrary requests to internal network resources. Public exploit code exists for this vulnerability, which poses significant risk in containerized environments where a compromised vLLM instance could be leveraged to access restricted internal systems. The vulnerability affects users running vLLM's multimodal features with untrusted input.

Denial Of Service Python Information Disclosure SSRF Vllm
NVD GitHub VulDB
CVSS 3.1
7.1
EPSS
0.0%
CVE-2026-22807 Jan 21, 22:15 PyPI CRITICAL PATCH GHSA Act Now

Remote code execution in vLLM 0.10.1 through 0.13.x lets an attacker who controls the model repository or path run arbitrary Python on the inference host: vLLM auto-loads Hugging Face `auto_map` dynamic modules during model resolution without honoring the `trust_remote_code` safety gate, so attacker-supplied code executes at server startup before any request is processed and without API access. The flaw carries a CVSS 9.8 and is fixed in 0.14.0, with Red Hat shipping multiple RHSA errata; however EPSS is only 0.06% (17th percentile) and there is no public exploit identified at time of analysis, so exploitation hinges on an attacker first influencing which model is loaded. This is a code-injection (CWE-94) trust-boundary bypass rather than a remotely reachable network service bug despite the AV:N rating.

RCE Python Code Injection Vllm
NVD GitHub VulDB
CVSS 3.1
9.8
EPSS
0.1%
CVE-2026-78684 Aug 25, 11:33 MEDIUM PATCH This Month

Remote denial-of-service in vLLM before 0.27.0 allows unauthenticated network attackers to activate the DeepStream GPU video-decode backend per-request, initialize the process-wide GPU decode pool, and submit video content that bypasses pixel-limit enforcement, causing partial availability loss for concurrent inference requests. The root cause is vLLM's failure to classify DeepStream as a GPU backend subject to resource controls combined with an omitted pixel-limit check in the DeepStream decode path (CWE-400). No public exploit code has been identified at time of analysis, and this vulnerability is not listed in the CISA KEV catalog.

Denial Of Service Red Hat Vllm Vllm Project
NVD GitHub VulDB
CVSS 4.0
6.9
EPSS
0.3%
CVE-2026-73560 Aug 17, 20:17 MEDIUM PATCH GHSA This Month

Server-side request forgery in vLLM's MiMoV2OmniMultiModalProcessor allows authenticated API users to bypass the framework's media access controls, enabling arbitrary file reads on the vLLM host and SSRF requests to internal network resources. The flaw affects all vLLM releases prior to 0.26.0 where attacker-controlled image and audio URL strings were passed directly to _fetch_image, requests.get, and Image.open instead of routing through the MediaConnector abstraction that enforces allowed_media_domains and allowed_local_media_path restrictions. No public exploit code or CISA KEV listing exists at time of analysis, but the low attack complexity and network-accessible vector make this a meaningful risk for deployments serving untrusted authenticated users via the affected multimodal processor.

SSRF Red Hat Vllm Vllm Project
NVD GitHub
CVSS 3.1
6.5
EPSS
0.3%
CVE-2026-73559 Aug 13, 15:09 PyPI MEDIUM POC PATCH GHSA This Month

Uncontrolled resource consumption in vLLM's OpenAI-compatible completions endpoint allows any authenticated API client to exhaust CPU, memory, async scheduling capacity, and engine request slots with a single crafted request. Affected versions span 0.19.0 through 0.25.x; the issue is fixed in 0.26.0. No public exploit code has been identified and CISA has not listed this in KEV, but the attack is trivially constructable by any client with a valid API credential.

Denial Of Service Vllm Vllm Project
NVD GitHub VulDB
CVSS 3.1
6.5
EPSS
0.4%
CVE-2026-73558 Aug 13, 15:06 MEDIUM POC PATCH GHSA This Month

Cross-user inference data leakage in vLLM prior to 0.27.0 exposes one batched user's prompt or completion output to another user sharing the same GPU inference batch. The root cause is a 32-bit integer overflow in the CUDA kernel `act_and_mul_kernel` within `activation_kernels.cu`, where the pointer arithmetic `blockIdx.x * 2 * d` wraps around when token indices and hidden-dimension sizes are large enough, redirecting the kernel to read from an adjacent user's input buffer. No active exploitation has been confirmed (no CISA KEV listing), but the vendor has patched the issue in v0.27.0 and a detailed GitHub advisory and PR diff are publicly available.

Buffer Overflow Integer Overflow Vllm Vllm Project
NVD GitHub VulDB
CVSS 3.1
5.3
EPSS
0.3%
CVE-2026-73557 Aug 13, 15:00 MEDIUM POC PATCH GHSA This Month

Race condition in vLLM's prompt embedding loader allows concurrent API requests to bypass the sparse tensor invariant guard introduced for CVE-2025-62164, permitting an invalid sparse tensor to reach `tensor.to_dense()` and cause availability disruption. Affected deployments span versions 0.20.2rc0 through 0.25.x and are limited to instances with `enable_prompt_embeds` explicitly enabled. No public exploit code or CISA KEV listing exists at time of analysis; the fix is confirmed in v0.26.0.

Information Disclosure Race Condition Vllm Vllm Project
NVD GitHub VulDB
CVSS 4.0
6.3
EPSS
0.3%
CVE-2026-73556 Aug 13, 14:56 MEDIUM POC PATCH GHSA This Month

Uncontrolled resource consumption in vLLM's lm-format-enforcer structured output backend (all versions prior to 0.26.0) enables unauthenticated remote denial of service via the /v1/completions API endpoint. By submitting a crafted catastrophically backtracking regular expression to the structured_outputs.regex parameter, a remote attacker can saturate an entire CPU core and stall the structured-output engine path indefinitely - blocking inference requests that rely on the lm-format-enforcer backend for all concurrent users. No public exploit code has been identified and the vulnerability is not listed in CISA KEV, though the attack primitive (ReDoS against an exposed API) is well-understood and trivially constructible by any attacker familiar with catastrophic regex patterns.

Denial Of Service Vllm Vllm Project
NVD GitHub VulDB
CVSS 3.1
5.3
EPSS
0.3%
CVE-2026-73555 Aug 13, 14:50 MEDIUM POC PATCH GHSA This Month

Information disclosure in vLLM prior to version 0.26.0 leaks server filesystem internals - including OS usernames, home and virtual-environment directory paths, Python version, internal package structure, line numbers, and endpoint handler names - through HTTP error responses. Any unauthenticated network attacker can trigger this by sending malformed JSON to the OpenAI-compatible API endpoints, requiring zero privileges and no special configuration beyond a default vLLM deployment. No public exploit code has been identified at time of analysis, and no CISA KEV listing exists; however, the trivial exploitation bar (a single malformed HTTP request) and the reconnaissance value of the leaked data elevate real-world risk above the CVSS 5.3 score alone implies.

Python Information Disclosure Vllm Vllm Project
NVD GitHub VulDB
CVSS 3.1
5.3
EPSS
0.3%
CVE-2026-55646 Jul 06, 19:41 PyPI MEDIUM POC PATCH GHSA This Month

Memory exhaustion in vLLM 0.22.0-0.23.0 allows authenticated API callers to crash or destabilize the inference server by uploading arbitrarily large audio files. The `/v1/audio/transcriptions` and `/v1/audio/translations` endpoints invoke `request.file.read()` to fully buffer multipart uploads into process memory before the `VLLM_MAX_AUDIO_CLIP_FILESIZE_MB` size guard is evaluated, meaning the size limit is checked only after the damage is done. No public exploit is identified at time of analysis; vendor-confirmed fix is available in version 0.24.0.

Denial Of Service Red Hat Vllm Vllm Project
NVD GitHub VulDB
CVSS 3.1
6.5
EPSS
0.3%
CVE-2025-71379 Jun 20, 18:27 PyPI MEDIUM PATCH This Month

Regular expression denial of service in vLLM versions 0.6.3 through 0.8.x exposes three distinct attack surfaces - the LoRA utility module, the phi4mini tool parser, and the OpenAI-compatible chat endpoint - to catastrophic regex backtracking, causing severe CPU exhaustion and service-wide denial of service. Authenticated API consumers can submit crafted inputs with deeply nested or repeated structures (e.g., `((((a|)+)+)+)`) to trigger unbounded processing in Python's backtracking NFA regex engine. No public exploit identified at time of analysis, though the GHSA advisory discloses the exact vulnerable patterns and example malicious inputs, substantially lowering the reproduction barrier for anyone with API access.

Denial Of Service Red Hat Vllm
NVD VulDB GitHub
CVSS 4.0
5.3
EPSS
0.2%
CVE-2026-9540 May 26, 10:30 PyPI MEDIUM POC PATCH This Month

Denial of service in vllm 0.19.0's OpenAI-compatible serving path allows remote unauthenticated attackers to exhaust scheduler resources by submitting requests with unbounded logprob counts. The root cause, confirmed by PR diff analysis, is the absence of any per-batch logprob budget in the v1 scheduler: requests specifying logprobs=-1 (full vocabulary) multiplied across parallel sequences (n) generate massive compute and memory overhead with no cap, blocking or crashing the inference server. Publicly available exploit code exists (GitHub issue #37343); no confirmed active exploitation at time of analysis.

Denial Of Service Vllm Vllm Project
NVD VulDB GitHub
CVSS 4.0
5.5
EPSS
0.1%
CVE-2026-7141 Apr 27, 17:16 PyPI LOW PATCH Monitor

Stale KV cache data leakage in vllm's KV Block Handler (versions 0.1-0.19.0) allows partial disclosure of prior inference requests' attention key/value data to subsequent requests. The `needs_kv_cache_zeroing` property incorrectly returned `False` for FullAttention model configurations - relying solely on `has_mamba_layers` - causing recycled KV cache blocks to go unzeroed between requests; partial-block tail slots retaining NaN/Inf values from prior sessions could then leak into the masked softmax computation of the next request. A publicly available PoC exists per SSVC and CVSS 4.0 E:P, though EPSS at 0.05% (14th percentile) and absence from CISA KEV indicate very low observed exploitation in practice.

Information Disclosure Vllm
NVD GitHub VulDB
CVSS 4.0
2.9
EPSS
0.0%
CVE-2026-22773 Jan 10, 07:16 PyPI MEDIUM POC PATCH This Month

Vllm versions up to 0.12.0 is affected by allocation of resources without limits or throttling (CVSS 6.5).

Denial Of Service Red Hat AI / ML Vllm
NVD GitHub
CVSS 3.1
6.5
EPSS
0.0%
CVE-2025-66448 Dec 01, 23:15 PyPI HIGH PATCH GHSA This Week

vLLM is an inference and serving engine for large language models (LLMs). Prior to 0.11.1, vllm has a critical remote code execution vector in a config class named Nemotron_Nano_VL_Config. When vllm loads a model config that contains an auto_map entry, the config class resolves that mapping with get_class_from_dynamic_module(...) and immediately instantiates the returned class. This fetches and executes Python from the remote repository referenced in the auto_map string. Crucially, this happens even when the caller explicitly sets trust_remote_code=False in vllm.transformers_utils.config.get_config. In practice, an attacker can publish a benign-looking frontend repo whose config.json points via auto_map to a separate malicious backend repo; loading the frontend will silently run the backend’s code on the victim host. This vulnerability is fixed in 0.11.1.

RCE Python Code Injection Debian Red Hat +1
NVD GitHub
CVSS 3.1
7.1
EPSS
0.2%
CVE-2025-62426 Nov 21, 02:15 PyPI MEDIUM PATCH This Month

vLLM is an inference and serving engine for large language models (LLMs). Rated medium severity (CVSS 6.5), this vulnerability is remotely exploitable, low attack complexity. This Allocation of Resources Without Limits vulnerability could allow attackers to exhaust system resources through uncontrolled allocation.

Denial Of Service Red Hat Vllm
NVD GitHub
CVSS 3.1
6.5
EPSS
0.1%
CVE-2025-62372 Nov 21, 02:15 PyPI HIGH PATCH GHSA This Month

vLLM is an inference and serving engine for large language models (LLMs). Rated high severity (CVSS 8.3), this vulnerability is remotely exploitable, low attack complexity.

Denial Of Service Red Hat Vllm
NVD GitHub
CVSS 4.0
8.3
EPSS
0.1%
CVE-2025-62164 Nov 21, 02:15 PyPI HIGH PATCH GHSA This Month

vLLM is an inference and serving engine for large language models (LLMs). Rated high severity (CVSS 8.8), this vulnerability is remotely exploitable, low attack complexity.

RCE Buffer Overflow Red Hat AI / ML Pytorch +2
NVD GitHub
CVSS 3.1
8.8
EPSS
0.1%
CVE-2025-48956 Aug 21, 15:15 PyPI HIGH PATCH GHSA This Month

vLLM is an inference and serving engine for large language models (LLMs). Rated high severity (CVSS 7.5), this vulnerability is remotely exploitable, no authentication required, low attack complexity. This Uncontrolled Resource Consumption vulnerability could allow attackers to cause denial of service by exhausting system resources.

Denial Of Service Red Hat Vllm
NVD GitHub
CVSS 3.1
7.5
EPSS
0.3%
CVE-2025-48944 May 30, 19:15 PyPI MEDIUM POC PATCH This Week

vLLM is an inference and serving engine for large language models (LLMs). Rated medium severity (CVSS 6.5), this vulnerability is remotely exploitable, low attack complexity. Public exploit code available and no vendor patch available.

Denial Of Service Red Hat Vllm
NVD GitHub
CVSS 3.1
6.5
EPSS
0.3%
CVE-2025-48943 May 30, 19:15 PyPI MEDIUM PATCH This Month

vLLM is an inference and serving engine for large language models (LLMs). Rated medium severity (CVSS 6.5), this vulnerability is remotely exploitable, low attack complexity.

Denial Of Service Red Hat Vllm
NVD GitHub
CVSS 3.1
6.5
EPSS
0.2%
CVE-2025-48942 May 30, 19:15 PyPI MEDIUM POC PATCH This Week

vLLM is an inference and serving engine for large language models (LLMs). Rated medium severity (CVSS 6.5), this vulnerability is remotely exploitable, low attack complexity. Public exploit code available.

Information Disclosure Red Hat Vllm
NVD GitHub
CVSS 3.1
6.5
EPSS
0.2%
CVE-2025-48887 May 30, 18:15 PyPI MEDIUM POC PATCH This Week

vLLM, an inference and serving engine for large language models (LLMs), has a Regular Expression Denial of Service (ReDoS) vulnerability in the file. Rated medium severity (CVSS 6.5), this vulnerability is remotely exploitable, low attack complexity. Public exploit code available.

Denial Of Service Red Hat Vllm
NVD GitHub
CVSS 3.1
6.5
EPSS
0.3%
CVE-2025-46722 May 29, 17:15 PyPI MEDIUM PATCH Monitor

vLLM is an inference and serving engine for large language models (LLMs). Rated medium severity (CVSS 4.2), this vulnerability is remotely exploitable.

Information Disclosure Red Hat Vllm
NVD GitHub
CVSS 3.1
4.2
EPSS
0.2%
CVE-2025-46570 May 29, 17:15 PyPI LOW PATCH Monitor

vLLM is an inference and serving engine for large language models (LLMs). Rated low severity (CVSS 2.6), this vulnerability is remotely exploitable.

Information Disclosure Vllm
NVD GitHub
CVSS 3.1
2.6
EPSS
0.2%
CVE-2025-47277 May 20, 18:15 PyPI CRITICAL POC PATCH GHSA Act Now

vLLM, an inference and serving engine for large language models (LLMs), has an issue in versions 0.6.5 through 0.8.4 that ONLY impacts environments using the `PyNcclPipe` KV cache transfer. Rated critical severity (CVSS 9.8), this vulnerability is remotely exploitable, no authentication required, low attack complexity. Public exploit code available.

Deserialization Red Hat AI / ML Pytorch Vllm
NVD GitHub
CVSS 3.1
9.8
EPSS
0.9%
CVE-2025-30165 May 06, 17:16 PyPI HIGH PATCH GHSA This Week

vLLM is an inference and serving engine for large language models. Rated high severity (CVSS 8.0), this vulnerability is low attack complexity. No vendor patch available.

RCE Deserialization Red Hat Vllm
NVD GitHub
CVSS 3.1
8.0
EPSS
1.3%
CVE-2025-46560 Apr 30, 01:15 PyPI MEDIUM POC PATCH This Month

vLLM is a high-throughput and memory-efficient inference and serving engine for LLMs. Rated medium severity (CVSS 6.5), this vulnerability is remotely exploitable, low attack complexity. Public exploit code available and no vendor patch available.

Denial Of Service Red Hat Vllm
NVD GitHub
CVSS 3.1
6.5
EPSS
0.6%
CVE-2025-32444 Apr 30, 01:15 PyPI CRITICAL POC PATCH Act Now

vLLM is a high-throughput and memory-efficient inference and serving engine for LLMs. Rated critical severity (CVSS 10.0), this vulnerability is remotely exploitable, no authentication required, low attack complexity. Public exploit code available.

RCE Deserialization Red Hat Vllm
NVD GitHub
CVSS 3.1
10.0
EPSS
2.5%
CVE-2025-30202 Apr 30, 01:15 PyPI HIGH POC PATCH GHSA This Week

vLLM is a high-throughput and memory-efficient inference and serving engine for LLMs. Rated high severity (CVSS 7.5), this vulnerability is remotely exploitable, no authentication required, low attack complexity. Public exploit code available.

Denial Of Service Information Disclosure Red Hat Vllm
NVD GitHub
CVSS 3.1
7.5
EPSS
0.4%
CVE-2024-9053 Mar 20, 10:15 PyPI CRITICAL POC GHSA Act Now

vllm-project vllm version 0.6.0 contains a vulnerability in the AsyncEngineRPCServer() RPC server entrypoints. Rated critical severity (CVSS 9.8), this vulnerability is remotely exploitable, no authentication required, low attack complexity. Public exploit code available and no vendor patch available.

RCE Deserialization Red Hat Vllm Vllm Project
NVD
CVSS 3.1
9.8
EPSS
2.2%
CVE-2024-11041 Mar 20, 10:15 PyPI CRITICAL POC Act Now

vllm-project vllm version v0.6.2 contains a vulnerability in the MessageQueue.dequeue() API function. Rated critical severity (CVSS 9.8), this vulnerability is remotely exploitable, no authentication required, low attack complexity. Public exploit code available and no vendor patch available.

RCE Deserialization Red Hat Vllm
NVD
CVSS 3.0
9.8
EPSS
1.3%
CVE-2025-29783 Mar 19, 16:15 PyPI CRITICAL PATCH Act Now

vLLM is a high-throughput and memory-efficient inference and serving engine for LLMs. Rated critical severity (CVSS 9.0), this vulnerability is low attack complexity. This Deserialization of Untrusted Data vulnerability could allow attackers to execute arbitrary code through malicious serialized objects.

RCE Deserialization Red Hat Vllm
NVD GitHub
CVSS 3.1
9.0
EPSS
1.7%
CVE-2025-29770 Mar 19, 16:15 PyPI MEDIUM PATCH This Month

vLLM is a high-throughput and memory-efficient inference and serving engine for LLMs. Rated medium severity (CVSS 6.5), this vulnerability is remotely exploitable, low attack complexity. This Allocation of Resources Without Limits vulnerability could allow attackers to exhaust system resources through uncontrolled allocation.

Denial Of Service Red Hat Vllm
NVD GitHub
CVSS 3.1
6.5
EPSS
0.3%
CVE-2025-25183 Feb 07, 20:15 PyPI LOW PATCH Monitor

vLLM is a high-throughput and memory-efficient inference and serving engine for LLMs. Rated low severity (CVSS 2.6), this vulnerability is remotely exploitable. No vendor patch available.

Python Information Disclosure Vllm
NVD GitHub
CVSS 3.1
2.6
EPSS
0.3%
CVE-2025-24357 Jan 27, 18:15 PyPI HIGH PATCH This Month

vLLM is a library for LLM inference and serving. Rated high severity (CVSS 7.5), this vulnerability is remotely exploitable, no authentication required. This Deserialization of Untrusted Data vulnerability could allow attackers to execute arbitrary code through malicious serialized objects.

RCE Deserialization Red Hat Vllm
NVD GitHub
CVSS 3.1
7.5
EPSS
1.0%
EPSS 0% CVSS 7.1
HIGH PATCH This Week

Denial of service in vLLM 0.12.0 through 0.23.x lets any authorized API caller crash the entire inference server by submitting a pure prompt-embeddings payload to the /v1/completions endpoint when a model using M-RoPE (multimodal rotary position embedding) is loaded. The malformed request trips a reachable assertion in the EngineCore process, which terminates the whole server rather than rejecting the single request. There is no public exploit identified at time of analysis and it is not listed in CISA KEV; the flaw is fixed in vLLM 0.24.0.

Denial Of Service Red Hat Vllm +1
NVD GitHub
EPSS 0% CVSS 8.7
HIGH PATCH This Week

Denial of service in vLLM inference servers prior to 0.24.0 allows remote unauthenticated attackers to hang an inference worker indefinitely by submitting a single request with an adversarial regular expression via the structured_outputs.regex API parameter. The pattern is passed to grammar compiler backends (xgrammar with no guard, outlines with structural-but-not-complexity validation) where nested quantifiers trigger exponential state-space expansion (ReDoS). No public exploit identified at time of analysis, though the trivial request-based trigger makes weaponization straightforward.

Denial Of Service Red Hat Vllm +1
NVD GitHub
EPSS 0% CVSS 7.5
HIGH POC PATCH This Week

Denial of service in the vLLM LLM inference server (all versions prior to 0.24.0) allows a remote client to crash the shared engine worker by sending a specific multi-request speculative decoding workload. The rejection sampler produces a recovered token equal to the vocabulary-size boundary, which is coerced to -1, written back into the drafter's input ids, and later dereferenced by the embedding/attention path, triggering a GPU device-side assertion that kills the worker. There is no public exploit identified at time of analysis and this CVE is not in CISA KEV; per CVSS the impact is availability-only (C:N/I:N/A:H) with a low EPSS profile expected for a crash-only bug.

Denial Of Service Red Hat Vllm +1
NVD GitHub
EPSS 0% CVSS 8.8
HIGH PATCH This Week

Remote code execution in vLLM versions prior to 0.22.1 allows attackers to backdoor production LLM inference deployments through a dependency confusion attack in the project's Dockerfile. Because flashinfer-jit-cache was pulled via --extra-index-url with UV_INDEX_STRATEGY=unsafe-best-match while the name remained unregistered on PyPI, any attacker who claimed the name on PyPI with a higher version would have their code executed as root during every Docker build. No public exploit identified at time of analysis, but the supply-chain primitive is well understood and trivially weaponizable.

RCE Docker Red Hat +2
NVD GitHub VulDB
EPSS 0% CVSS 8.7
HIGH PATCH This Week

Denial of service and potential memory corruption in vLLM versions 0.10.2 through 0.12.x stems from missing sparse tensor validation in multimodal embeddings processing, allowing authenticated remote users to submit crafted prompt-embedding requests with malformed tensor indices. Because PyTorch disables sparse tensor invariant checks by default, attackers can crash the inference server or exhaust resources, with potential out-of-bounds or write-what-where memory corruption. No public exploit identified at time of analysis; this continuation of CVE-2025-62164 addresses the root cause that the prior fix only masked by disabling the feature by default.

Buffer Overflow Denial Of Service Red Hat +1
NVD GitHub VulDB
EPSS 0% CVSS 7.1
HIGH PATCH This Week

vLLM versions 0.5.5 through 0.17.x use incorrect mono audio downmixing via numpy.mean instead of the ITU-R BS.775-4 weighted standard, causing audio processed by AI models to diverge from human perception. An authenticated remote attacker with low privileges can exploit this inconsistency to manipulate audio-based model outputs or infer mismatches between expected and actual audio processing, affecting integrity of audio-driven inference pipelines. The vulnerability has been patched in vLLM 0.18.0.

Information Disclosure Vllm Vllm Project
NVD GitHub VulDB
EPSS 0% CVSS 8.8
HIGH PATCH This Week

Remote code execution is possible in vLLM inference and serving engine versions 0.10.1 through 0.17.x due to hardcoded trust_remote_code=True settings in two model implementation files that override users' explicit --trust-remote-code=False security configuration. Attackers can exploit this by hosting malicious model repositories that execute arbitrary code when loaded by vLLM, even when users have intentionally disabled remote code trust for security. Version 0.18.0 patches this vulnerability, with no public exploit identified at time of analysis and a CVSS score of 8.8 requiring user interaction to trigger.

RCE Vllm
NVD GitHub VulDB
EPSS 0% CVSS 9.8
CRITICAL PATCH Act Now

Server-Side Request Forgery in vLLM's multimodal MediaConnector allows remote attackers to coerce the inference server into fetching attacker-chosen internal URLs, bypassing the allowed_media_domains allowlist that was added to fix CVE-2026-24779. The bypass exploits a backslash-before-@ parsing disagreement between urllib3.util.parse_url (validation layer) and aiohttp/yarl (HTTP client), so a URL the allowlist reads as a permitted host is actually fetched from an internal target. It carries a CVSS 9.8 rating, a vendor patch and Red Hat advisory are available, and a working bypass payload is published in the GHSA advisory and fix test suite, though EPSS is very low (0.02%, 3rd percentile) and it is not on CISA KEV.

SSRF Vllm
NVD GitHub VulDB
EPSS 0% CVSS 9.8
CRITICAL POC PATCH Act Now

Information exposure in vLLM inference engine versions 0.8.3 to before 0.14.1. Invalid image requests to the multimodal endpoint cause sensitive data logging. Patch available.

RCE Vllm
NVD GitHub VulDB
EPSS 0% CVSS 7.1
HIGH POC PATCH This Week

vLLM before version 0.14.1 contains a server-side request forgery vulnerability in the MediaConnector class where inconsistent URL parsing between libraries allows attackers to bypass host restrictions and force the server to make arbitrary requests to internal network resources. Public exploit code exists for this vulnerability, which poses significant risk in containerized environments where a compromised vLLM instance could be leveraged to access restricted internal systems. The vulnerability affects users running vLLM's multimodal features with untrusted input.

Denial Of Service Python Information Disclosure +2
NVD GitHub VulDB
EPSS 0% CVSS 9.8
CRITICAL PATCH Act Now

Remote code execution in vLLM 0.10.1 through 0.13.x lets an attacker who controls the model repository or path run arbitrary Python on the inference host: vLLM auto-loads Hugging Face `auto_map` dynamic modules during model resolution without honoring the `trust_remote_code` safety gate, so attacker-supplied code executes at server startup before any request is processed and without API access. The flaw carries a CVSS 9.8 and is fixed in 0.14.0, with Red Hat shipping multiple RHSA errata; however EPSS is only 0.06% (17th percentile) and there is no public exploit identified at time of analysis, so exploitation hinges on an attacker first influencing which model is loaded. This is a code-injection (CWE-94) trust-boundary bypass rather than a remotely reachable network service bug despite the AV:N rating.

RCE Python Code Injection +1
NVD GitHub VulDB
EPSS 0% CVSS 6.9
MEDIUM PATCH This Month

Remote denial-of-service in vLLM before 0.27.0 allows unauthenticated network attackers to activate the DeepStream GPU video-decode backend per-request, initialize the process-wide GPU decode pool, and submit video content that bypasses pixel-limit enforcement, causing partial availability loss for concurrent inference requests. The root cause is vLLM's failure to classify DeepStream as a GPU backend subject to resource controls combined with an omitted pixel-limit check in the DeepStream decode path (CWE-400). No public exploit code has been identified at time of analysis, and this vulnerability is not listed in the CISA KEV catalog.

Denial Of Service Red Hat Vllm +1
NVD GitHub VulDB
EPSS 0% CVSS 6.5
MEDIUM PATCH This Month

Server-side request forgery in vLLM's MiMoV2OmniMultiModalProcessor allows authenticated API users to bypass the framework's media access controls, enabling arbitrary file reads on the vLLM host and SSRF requests to internal network resources. The flaw affects all vLLM releases prior to 0.26.0 where attacker-controlled image and audio URL strings were passed directly to _fetch_image, requests.get, and Image.open instead of routing through the MediaConnector abstraction that enforces allowed_media_domains and allowed_local_media_path restrictions. No public exploit code or CISA KEV listing exists at time of analysis, but the low attack complexity and network-accessible vector make this a meaningful risk for deployments serving untrusted authenticated users via the affected multimodal processor.

SSRF Red Hat Vllm +1
NVD GitHub
EPSS 0% CVSS 6.5
MEDIUM POC PATCH This Month

Uncontrolled resource consumption in vLLM's OpenAI-compatible completions endpoint allows any authenticated API client to exhaust CPU, memory, async scheduling capacity, and engine request slots with a single crafted request. Affected versions span 0.19.0 through 0.25.x; the issue is fixed in 0.26.0. No public exploit code has been identified and CISA has not listed this in KEV, but the attack is trivially constructable by any client with a valid API credential.

Denial Of Service Vllm Vllm Project
NVD GitHub VulDB
EPSS 0% CVSS 5.3
MEDIUM POC PATCH This Month

Cross-user inference data leakage in vLLM prior to 0.27.0 exposes one batched user's prompt or completion output to another user sharing the same GPU inference batch. The root cause is a 32-bit integer overflow in the CUDA kernel `act_and_mul_kernel` within `activation_kernels.cu`, where the pointer arithmetic `blockIdx.x * 2 * d` wraps around when token indices and hidden-dimension sizes are large enough, redirecting the kernel to read from an adjacent user's input buffer. No active exploitation has been confirmed (no CISA KEV listing), but the vendor has patched the issue in v0.27.0 and a detailed GitHub advisory and PR diff are publicly available.

Buffer Overflow Integer Overflow Vllm +1
NVD GitHub VulDB
EPSS 0% CVSS 6.3
MEDIUM POC PATCH This Month

Race condition in vLLM's prompt embedding loader allows concurrent API requests to bypass the sparse tensor invariant guard introduced for CVE-2025-62164, permitting an invalid sparse tensor to reach `tensor.to_dense()` and cause availability disruption. Affected deployments span versions 0.20.2rc0 through 0.25.x and are limited to instances with `enable_prompt_embeds` explicitly enabled. No public exploit code or CISA KEV listing exists at time of analysis; the fix is confirmed in v0.26.0.

Information Disclosure Race Condition Vllm +1
NVD GitHub VulDB
EPSS 0% CVSS 5.3
MEDIUM POC PATCH This Month

Uncontrolled resource consumption in vLLM's lm-format-enforcer structured output backend (all versions prior to 0.26.0) enables unauthenticated remote denial of service via the /v1/completions API endpoint. By submitting a crafted catastrophically backtracking regular expression to the structured_outputs.regex parameter, a remote attacker can saturate an entire CPU core and stall the structured-output engine path indefinitely - blocking inference requests that rely on the lm-format-enforcer backend for all concurrent users. No public exploit code has been identified and the vulnerability is not listed in CISA KEV, though the attack primitive (ReDoS against an exposed API) is well-understood and trivially constructible by any attacker familiar with catastrophic regex patterns.

Denial Of Service Vllm Vllm Project
NVD GitHub VulDB
EPSS 0% CVSS 5.3
MEDIUM POC PATCH This Month

Information disclosure in vLLM prior to version 0.26.0 leaks server filesystem internals - including OS usernames, home and virtual-environment directory paths, Python version, internal package structure, line numbers, and endpoint handler names - through HTTP error responses. Any unauthenticated network attacker can trigger this by sending malformed JSON to the OpenAI-compatible API endpoints, requiring zero privileges and no special configuration beyond a default vLLM deployment. No public exploit code has been identified at time of analysis, and no CISA KEV listing exists; however, the trivial exploitation bar (a single malformed HTTP request) and the reconnaissance value of the leaked data elevate real-world risk above the CVSS 5.3 score alone implies.

Python Information Disclosure Vllm +1
NVD GitHub VulDB
EPSS 0% CVSS 6.5
MEDIUM POC PATCH This Month

Memory exhaustion in vLLM 0.22.0-0.23.0 allows authenticated API callers to crash or destabilize the inference server by uploading arbitrarily large audio files. The `/v1/audio/transcriptions` and `/v1/audio/translations` endpoints invoke `request.file.read()` to fully buffer multipart uploads into process memory before the `VLLM_MAX_AUDIO_CLIP_FILESIZE_MB` size guard is evaluated, meaning the size limit is checked only after the damage is done. No public exploit is identified at time of analysis; vendor-confirmed fix is available in version 0.24.0.

Denial Of Service Red Hat Vllm +1
NVD GitHub VulDB
EPSS 0% CVSS 5.3
MEDIUM PATCH This Month

Regular expression denial of service in vLLM versions 0.6.3 through 0.8.x exposes three distinct attack surfaces - the LoRA utility module, the phi4mini tool parser, and the OpenAI-compatible chat endpoint - to catastrophic regex backtracking, causing severe CPU exhaustion and service-wide denial of service. Authenticated API consumers can submit crafted inputs with deeply nested or repeated structures (e.g., `((((a|)+)+)+)`) to trigger unbounded processing in Python's backtracking NFA regex engine. No public exploit identified at time of analysis, though the GHSA advisory discloses the exact vulnerable patterns and example malicious inputs, substantially lowering the reproduction barrier for anyone with API access.

Denial Of Service Red Hat Vllm
NVD VulDB GitHub
EPSS 0% CVSS 5.5
MEDIUM POC PATCH This Month

Denial of service in vllm 0.19.0's OpenAI-compatible serving path allows remote unauthenticated attackers to exhaust scheduler resources by submitting requests with unbounded logprob counts. The root cause, confirmed by PR diff analysis, is the absence of any per-batch logprob budget in the v1 scheduler: requests specifying logprobs=-1 (full vocabulary) multiplied across parallel sequences (n) generate massive compute and memory overhead with no cap, blocking or crashing the inference server. Publicly available exploit code exists (GitHub issue #37343); no confirmed active exploitation at time of analysis.

Denial Of Service Vllm Vllm Project
NVD VulDB GitHub
EPSS 0% CVSS 2.9
LOW PATCH Monitor

Stale KV cache data leakage in vllm's KV Block Handler (versions 0.1-0.19.0) allows partial disclosure of prior inference requests' attention key/value data to subsequent requests. The `needs_kv_cache_zeroing` property incorrectly returned `False` for FullAttention model configurations - relying solely on `has_mamba_layers` - causing recycled KV cache blocks to go unzeroed between requests; partial-block tail slots retaining NaN/Inf values from prior sessions could then leak into the masked softmax computation of the next request. A publicly available PoC exists per SSVC and CVSS 4.0 E:P, though EPSS at 0.05% (14th percentile) and absence from CISA KEV indicate very low observed exploitation in practice.

Information Disclosure Vllm
NVD GitHub VulDB
EPSS 0% CVSS 6.5
MEDIUM POC PATCH This Month

Vllm versions up to 0.12.0 is affected by allocation of resources without limits or throttling (CVSS 6.5).

Denial Of Service Red Hat AI / ML +1
NVD GitHub
EPSS 0% CVSS 7.1
HIGH PATCH This Week

vLLM is an inference and serving engine for large language models (LLMs). Prior to 0.11.1, vllm has a critical remote code execution vector in a config class named Nemotron_Nano_VL_Config. When vllm loads a model config that contains an auto_map entry, the config class resolves that mapping with get_class_from_dynamic_module(...) and immediately instantiates the returned class. This fetches and executes Python from the remote repository referenced in the auto_map string. Crucially, this happens even when the caller explicitly sets trust_remote_code=False in vllm.transformers_utils.config.get_config. In practice, an attacker can publish a benign-looking frontend repo whose config.json points via auto_map to a separate malicious backend repo; loading the frontend will silently run the backend’s code on the victim host. This vulnerability is fixed in 0.11.1.

RCE Python Code Injection +3
NVD GitHub
EPSS 0% CVSS 6.5
MEDIUM PATCH This Month

vLLM is an inference and serving engine for large language models (LLMs). Rated medium severity (CVSS 6.5), this vulnerability is remotely exploitable, low attack complexity. This Allocation of Resources Without Limits vulnerability could allow attackers to exhaust system resources through uncontrolled allocation.

Denial Of Service Red Hat Vllm
NVD GitHub
EPSS 0% CVSS 8.3
HIGH PATCH This Month

vLLM is an inference and serving engine for large language models (LLMs). Rated high severity (CVSS 8.3), this vulnerability is remotely exploitable, low attack complexity.

Denial Of Service Red Hat Vllm
NVD GitHub
EPSS 0% CVSS 8.8
HIGH PATCH This Month

vLLM is an inference and serving engine for large language models (LLMs). Rated high severity (CVSS 8.8), this vulnerability is remotely exploitable, low attack complexity.

RCE Buffer Overflow Red Hat +4
NVD GitHub
EPSS 0% CVSS 7.5
HIGH PATCH This Month

vLLM is an inference and serving engine for large language models (LLMs). Rated high severity (CVSS 7.5), this vulnerability is remotely exploitable, no authentication required, low attack complexity. This Uncontrolled Resource Consumption vulnerability could allow attackers to cause denial of service by exhausting system resources.

Denial Of Service Red Hat Vllm
NVD GitHub
EPSS 0% CVSS 6.5
MEDIUM POC PATCH This Week

vLLM is an inference and serving engine for large language models (LLMs). Rated medium severity (CVSS 6.5), this vulnerability is remotely exploitable, low attack complexity. Public exploit code available and no vendor patch available.

Denial Of Service Red Hat Vllm
NVD GitHub
EPSS 0% CVSS 6.5
MEDIUM PATCH This Month

vLLM is an inference and serving engine for large language models (LLMs). Rated medium severity (CVSS 6.5), this vulnerability is remotely exploitable, low attack complexity.

Denial Of Service Red Hat Vllm
NVD GitHub
EPSS 0% CVSS 6.5
MEDIUM POC PATCH This Week

vLLM is an inference and serving engine for large language models (LLMs). Rated medium severity (CVSS 6.5), this vulnerability is remotely exploitable, low attack complexity. Public exploit code available.

Information Disclosure Red Hat Vllm
NVD GitHub
EPSS 0% CVSS 6.5
MEDIUM POC PATCH This Week

vLLM, an inference and serving engine for large language models (LLMs), has a Regular Expression Denial of Service (ReDoS) vulnerability in the file. Rated medium severity (CVSS 6.5), this vulnerability is remotely exploitable, low attack complexity. Public exploit code available.

Denial Of Service Red Hat Vllm
NVD GitHub
EPSS 0% CVSS 4.2
MEDIUM PATCH Monitor

vLLM is an inference and serving engine for large language models (LLMs). Rated medium severity (CVSS 4.2), this vulnerability is remotely exploitable.

Information Disclosure Red Hat Vllm
NVD GitHub
EPSS 0% CVSS 2.6
LOW PATCH Monitor

vLLM is an inference and serving engine for large language models (LLMs). Rated low severity (CVSS 2.6), this vulnerability is remotely exploitable.

Information Disclosure Vllm
NVD GitHub
EPSS 1% CVSS 9.8
CRITICAL POC PATCH Act Now

vLLM, an inference and serving engine for large language models (LLMs), has an issue in versions 0.6.5 through 0.8.4 that ONLY impacts environments using the `PyNcclPipe` KV cache transfer. Rated critical severity (CVSS 9.8), this vulnerability is remotely exploitable, no authentication required, low attack complexity. Public exploit code available.

Deserialization Red Hat AI / ML +2
NVD GitHub
EPSS 1% CVSS 8.0
HIGH PATCH This Week

vLLM is an inference and serving engine for large language models. Rated high severity (CVSS 8.0), this vulnerability is low attack complexity. No vendor patch available.

RCE Deserialization Red Hat +1
NVD GitHub
EPSS 1% CVSS 6.5
MEDIUM POC PATCH This Month

vLLM is a high-throughput and memory-efficient inference and serving engine for LLMs. Rated medium severity (CVSS 6.5), this vulnerability is remotely exploitable, low attack complexity. Public exploit code available and no vendor patch available.

Denial Of Service Red Hat Vllm
NVD GitHub
EPSS 2% CVSS 10.0
CRITICAL POC PATCH Act Now

vLLM is a high-throughput and memory-efficient inference and serving engine for LLMs. Rated critical severity (CVSS 10.0), this vulnerability is remotely exploitable, no authentication required, low attack complexity. Public exploit code available.

RCE Deserialization Red Hat +1
NVD GitHub
EPSS 0% CVSS 7.5
HIGH POC PATCH This Week

vLLM is a high-throughput and memory-efficient inference and serving engine for LLMs. Rated high severity (CVSS 7.5), this vulnerability is remotely exploitable, no authentication required, low attack complexity. Public exploit code available.

Denial Of Service Information Disclosure Red Hat +1
NVD GitHub
EPSS 2% CVSS 9.8
CRITICAL POC Act Now

vllm-project vllm version 0.6.0 contains a vulnerability in the AsyncEngineRPCServer() RPC server entrypoints. Rated critical severity (CVSS 9.8), this vulnerability is remotely exploitable, no authentication required, low attack complexity. Public exploit code available and no vendor patch available.

RCE Deserialization Red Hat +2
NVD
EPSS 1% CVSS 9.8
CRITICAL POC Act Now

vllm-project vllm version v0.6.2 contains a vulnerability in the MessageQueue.dequeue() API function. Rated critical severity (CVSS 9.8), this vulnerability is remotely exploitable, no authentication required, low attack complexity. Public exploit code available and no vendor patch available.

RCE Deserialization Red Hat +1
NVD
EPSS 2% CVSS 9.0
CRITICAL PATCH Act Now

vLLM is a high-throughput and memory-efficient inference and serving engine for LLMs. Rated critical severity (CVSS 9.0), this vulnerability is low attack complexity. This Deserialization of Untrusted Data vulnerability could allow attackers to execute arbitrary code through malicious serialized objects.

RCE Deserialization Red Hat +1
NVD GitHub
EPSS 0% CVSS 6.5
MEDIUM PATCH This Month

vLLM is a high-throughput and memory-efficient inference and serving engine for LLMs. Rated medium severity (CVSS 6.5), this vulnerability is remotely exploitable, low attack complexity. This Allocation of Resources Without Limits vulnerability could allow attackers to exhaust system resources through uncontrolled allocation.

Denial Of Service Red Hat Vllm
NVD GitHub
EPSS 0% CVSS 2.6
LOW PATCH Monitor

vLLM is a high-throughput and memory-efficient inference and serving engine for LLMs. Rated low severity (CVSS 2.6), this vulnerability is remotely exploitable. No vendor patch available.

Python Information Disclosure Vllm
NVD GitHub
EPSS 1% CVSS 7.5
HIGH PATCH This Month

vLLM is a library for LLM inference and serving. Rated high severity (CVSS 7.5), this vulnerability is remotely exploitable, no authentication required. This Deserialization of Untrusted Data vulnerability could allow attackers to execute arbitrary code through malicious serialized objects.

RCE Deserialization Red Hat +1
NVD GitHub

This site uses cookies essential for authentication and security. No tracking or analytics cookies are used. Privacy Policy