Vulnerability database
Every flaw the U.S. government’s cybersecurity agency (CISA) has confirmed attackers are using, plus the new WordPress plugin and theme bugs published each day, sorted by what they mean for a small business. We refresh it twice a day. If you see something your website or office runs, tell us.
136 results · page 1 of 3
Oct 26, 2026vLLM is an inference and serving engine for large language models.CVE-2026-115918 · Other software · vllm-project · reported by GitHubMedium 6.5
vLLM is an inference and serving engine for large language models. From 0.16.0 until 0.32.0, the /v1/responses endpoint allows an authenticated caller to place an attacker-controlled template in chat_template_kwargs[chat_template] when no server template is configured. ChatParams.get_apply_chat_template_kwargs() preserves that value without enforcing trust_request_chat_template, so the request can bypass the intended trust policy and cause CPU or memory exhaustion through Jinja rendering. This issue is fixed in version 0.32.0.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Oct 26, 2026vLLM is an inference and serving engine for large language models.CVE-2026-115917 · Other software · vllm-project · reported by GitHubLow 3.4
vLLM is an inference and serving engine for large language models. Prior to 0.32.0, the OpenAI-compatible server configures Starlette CORSMiddleware with allowed_origins set to a wildcard by default. A malicious web page can use a victim's browser to reach an unauthenticated vLLM server on localhost, an internal network, or a VPN, call endpoints including GET /v1/models and POST /v1/chat/completions, consume GPU resources, and read responses. Cookie-based cross-site request forgery does not apply because credentials are disabled by default, and deployments requiring bearer authentication are exposed only when the caller already has a usable token. This issue is fixed in version 0.32.0.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Oct 26, 2026vLLM is an inference and serving engine for large language models.CVE-2026-115916 · Other software · vllm-project · reported by GitHubMedium 6.5
vLLM is an inference and serving engine for large language models. Prior to 0.32.0, vllm/tokenizers/hf.py computes max_token_id with an off-by-one upper bound, allowing a pre-tokenized prompt ID equal to the vocabulary size to pass validation. For a tensor-parallel size of one and an exact multiple-of-64 vocabulary, VocabParallelEmbedding.forward does not mask that value before F.embedding uses it, causing a CUDA device-side assertion and terminating the engine. This issue is fixed in version 0.32.0.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Oct 26, 2026vLLM is an inference and serving engine for large language models.CVE-2026-115915 · Other software · vllm-project · reported by GitHubHigh 8.1
vLLM is an inference and serving engine for large language models. Prior to 0.32.0, ImageMediaIO passes attacker-controlled image bytes to Pillow through Image.open(BytesIO(data)) without restricting decoder selection, and the MIME label does not constrain the selected format. Image-capable request paths through parse_image() and fetch_image() can therefore dispatch an EPS payload to host components. A host with Ghostscript affected by CVE-2024-29510 can execute commands with the vLLM process's privileges, while environments using Pillow 12.0.0 through 12.2.0 can trigger CVE-2026-59203 and exhaust media-loading workers. Deployments without the applicable vulnerable dependency are not affected by that corresponding impact. This issue is fixed in version 0.32.0.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Oct 26, 2026vLLM is an inference and serving engine for large language models.CVE-2026-115914 · Other software · vllm-project · reported by GitHubMedium 6.5
vLLM is an inference and serving engine for large language models. From 0.11.1 until 0.32.0, the DeepSeek-OCR, DeepSeek-OCR-2, and Unlimited-OCR image path allows a remote client that can reach the OpenAI-compatible API to submit a legal image whose normalized pixels sum to zero when that image is alone in its per-step encoder batch. _parse_and_validate_image_input returns no embedding, embed_multimodal propagates that value, and sanity_check_mm_encoder_outputs triggers a fatal assertion that shuts down the engine for all users. This issue is fixed in version 0.32.0.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Oct 26, 2026vLLM is an inference and serving engine for large language models.CVE-2026-115913 · Other software · vllm-project · reported by GitHubMedium 6.5
vLLM is an inference and serving engine for large language models. Prior to 0.32.0, the Phi-4 multimodal audio encoder in phi4mm_audio.py and phi4mm_utils.py uses eager attention by default, allowing a remote client to submit one legal long audio clip that causes MultiHeadedAttention.forward_attention to materialize and retain full attention matrices in self.attn. The scaled dot-product attention branch does not materialize these matrices. The resulting quadratic GPU-memory use can cause an out-of-memory condition and terminate EngineCore for all users. This issue is fixed in version 0.32.0.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Oct 26, 2026vLLM is an inference and serving engine for large language models.CVE-2026-115912 · Other software · vllm-project · reported by GitHubHigh 7.5
vLLM is an inference and serving engine for large language models. Prior to 0.32.0, an authenticated caller can place forged shared-memory address and monotonic_id values in public multimodal features.kwargs_data. In scale-out deployments with shared-memory multimodal caching enabled, workers trust those values and pass attacker-selected bytes to pickle.loads(), allowing arbitrary Python execution even when insecure serialization is disabled. This issue is fixed in version 0.32.0.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Oct 26, 2026vLLM is an inference and serving engine for large language models.CVE-2026-115911 · Other software · vllm-project · reported by GitHubMedium 6.5
vLLM is an inference and serving engine for large language models. From 0.19.0 until 0.32.0, authenticated callers can supply a request chat template when no server-side --chat-template is configured, or a malicious model repository can provide a template that is compiled and rendered by tokenizer.apply_chat_template() without a wall-clock timeout, output-size limit, or iteration bound. Rendering runs in a ThreadPoolExecutor with one worker by default, so a nonterminating render can block /v1/responses and other chat-template users after the request ends, while an output-expanding render can exhaust server memory. Harmony models are not affected because they do not use this Jinja path. This issue is fixed in version 0.32.0.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Oct 25, 2026vLLM from 0.20.1 before 0.29.0 contains an improper input validation vulnerability that allows remote attackers to crash the engine by supplying out-of-vocabulaCVE-2026-116181 · Other software · vllm-project · reported by VulnCheckHigh 7.1
vLLM from 0.20.1 before 0.29.0 contains an improper input validation vulnerability that allows remote attackers to crash the engine by supplying out-of-vocabulary stop_token_ids. Attackers can send OpenAI chat, completion, generate, or batch requests with invalid IDs to terminate the engine or write outside the V2 runner's logits buffer.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD recordgithub.comvulncheck.com
Oct 25, 2026vLLM from 0.8.5 before 0.31.0 has a denial of service flaw.CVE-2026-116180 · Other software · vllm-project · reported by VulnCheckHigh 8.3
vLLM from 0.8.5 before 0.31.0 has a denial of service flaw. LogprobsTensors.empty_cpu() in vllm/v1/outputs.py allocates an uninitialized int32 tensor, and the V1 model runner leaves prompt positions reported as computed by a KV connector unwritten. A client sharing the engine can send prompt_logprobs (or echo with logprobs) against a connector-cached prefix, making tokenizer decode raise OverflowError, which crashes EngineCore and drops all co-tenant requests.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD recordgithub.comvulncheck.com
Oct 25, 2026vLLM from 0.18.0 before 0.31.0 contains a resource exhaustion vulnerability that allows API clients to bypass the video/jpeg frame-count cap by overriding num_fCVE-2026-116179 · Other software · vllm-project · reported by VulnCheckHigh 8.7
vLLM from 0.18.0 before 0.31.0 contains a resource exhaustion vulnerability that allows API clients to bypass the video/jpeg frame-count cap by overriding num_frames through media_io_kwargs. Attackers can set num_frames to -1 with a large inline video so VideoMediaIO.load_base64 decodes every frame, exhausting memory and crashing the API frontend.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD recordgithub.comvulncheck.com
Oct 25, 2026vLLM 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 meCVE-2026-116178 · Other software · vllm-project · reported by VulnCheckHigh 8.2
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 mm_placeholders ranges or mismatched kwargs_data and mm_hashes lists, producing a negative text span in Qwen2.5-VL position construction that terminates the engine.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD recordgithub.comvulncheck.com
Oct 25, 2026vLLM from 0.10.2 before 0.25.0 contains an uncontrolled resource consumption vulnerability that allows unauthenticated attackers to exhaust memory because the NCVE-2026-116177 · Other software · vllm-project · reported by VulnCheckHigh 8.7
vLLM from 0.10.2 before 0.25.0 contains an uncontrolled resource consumption vulnerability that allows unauthenticated attackers to exhaust memory because the Nemotron-VL processors globally set Image.MAX_IMAGE_PIXELS to None. Attackers can submit a pathological decompression-bomb image to /v1/chat/completions to OOM-kill the worker and deny service to co-tenants.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD recordgithub.comvulncheck.com
Oct 25, 2026vLLM from 0.9.2 before 0.24.0 contains a server-side request forgery vulnerability that allows API callers to force outbound connections by embedding a decode aCVE-2026-116176 · Other software · vllm-project · reported by VulnCheckHigh 8.8
vLLM from 0.9.2 before 0.24.0 contains a server-side request forgery vulnerability that allows API callers to force outbound connections by embedding a decode address in the X-Request-Id header. Attackers can make a P2pNcclConnector prefill worker ZMQ-connect to an attacker host, leaking the NCCL unique_id and KV-cache tensors or blocking the worker.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD recordgithub.comvulncheck.com
Oct 25, 2026vLLM from 0.19.0 before 0.24.0 contains a decompression bomb vulnerability because the AudioMediaIO multimodal path calls load_audio without the max_duration_s CVE-2026-116175 · Other software · vllm-project · reported by VulnCheckHigh 7.1
vLLM from 0.19.0 before 0.24.0 contains a decompression bomb vulnerability because the AudioMediaIO multimodal path calls load_audio without the max_duration_s decoded-duration limit. Authenticated attackers can send small compressed audio via input_audio or audio_url in /v1/chat/completions requests that expands to huge decoded PCM, exhausting server memory.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD recordgithub.comvulncheck.com
Oct 25, 2026vLLM from 0.22.0 before 0.24.0 contains an improper input validation vulnerability in the Rust gRPC Generate and GenerateStream service that forwards unchecked CVE-2026-116174 · Other software · vllm-project · reported by VulnCheckHigh 8.2
vLLM from 0.22.0 before 0.24.0 contains an improper input validation vulnerability in the Rust gRPC Generate and GenerateStream service that forwards unchecked prompt token IDs to EngineCore. Attackers can submit requests with out-of-vocabulary prompt.token_ids values to trigger a fatal embedding lookup failure that terminates EngineCore and leaves the server unhealthy.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD recordgithub.comvulncheck.com
Oct 25, 2026vLLM from 0.22.0 before 0.24.0 contains a denial of service vulnerability that allows remote attackers to terminate EngineCore by bypassing logprob_token_ids liCVE-2026-116173 · Other software · vllm-project · reported by VulnCheckHigh 8.2
vLLM from 0.22.0 before 0.24.0 contains a denial of service vulnerability that allows remote attackers to terminate EngineCore by bypassing logprob_token_ids limits in the Rust gRPC frontend. Attackers can send Generate or GenerateStream requests with output_logprobs enabled and more than 128 output_candidates.token_ids, triggering a fatal sampler exception that crashes the engine.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD recordgithub.comvulncheck.com
Oct 25, 2026vLLM from 0.22.0 before 0.24.0 contains an improper input validation vulnerability in the Rust gRPC converter that allows remote attackers to bypass max_logprobCVE-2026-116172 · Other software · vllm-project · reported by VulnCheckHigh 8.2
vLLM from 0.22.0 before 0.24.0 contains an improper input validation vulnerability in the Rust gRPC converter that allows remote attackers to bypass max_logprobs limits via CandidateTokens.top_n. Attackers can send Generate or GenerateStream requests with oversized or integer-wrapping top_n values to trigger torch.topk failures that terminate EngineCore.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD recordgithub.comvulncheck.com
Oct 20, 2026vLLM from 0.7.3 before 0.31.0 contains an uncontrolled resource consumption vulnerability that allows remote clients to control multimodal processor parameters CVE-2026-114055 · Other software · vllm-project · reported by VulnCheckHigh 8.7
vLLM from 0.7.3 before 0.31.0 contains an uncontrolled resource consumption vulnerability that allows remote clients to control multimodal processor parameters by supplying per-request mm_processor_kwargs. Attackers can set large image size, patch count, crop count, or frame count values in Isaac, InternVL, H2OVL, and other processors to exhaust memory and terminate the shared server.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD recordgithub.comvulncheck.com
Oct 20, 2026vLLM before 0.31.0 contains a cache aliasing vulnerability that allows authenticated attackers to collide shared cache keys by supplying caller-controlled valueCVE-2026-114054 · Other software · vllm-project · reported by VulnCheckHigh 7.1
vLLM before 0.31.0 contains a cache aliasing vulnerability that allows authenticated attackers to collide shared cache keys by supplying caller-controlled values such as EXIF ImageID or cache_salt. Attackers can set a UUID EXIF ImageID to receive another request's cached processed media, or match cache_salt to a LoRA name to reuse first-block KV state.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD recordgithub.comvulncheck.com
Oct 20, 2026vLLM from 0.7.3 before 0.31.0 forwards the request-supplied mm_processor_kwargs.code_revision to Transformers AutoProcessor.from_pretrained, letting clients selCVE-2026-114053 · Other software · vllm-project · reported by VulnCheckCritical 9.2
vLLM from 0.7.3 before 0.31.0 forwards the request-supplied mm_processor_kwargs.code_revision to Transformers AutoProcessor.from_pretrained, letting clients select which revision of a model's dynamic processor code is imported when trust_remote_code is enabled. An attacker controlling a revision in the served model repository can send a multimodal request to the unauthenticated /tokenize endpoint to execute arbitrary Python in the vLLM API process.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Oct 20, 2026vLLM after 0.24.0 and before 0.31.0 contains a memory exhaustion vulnerability in MOSS-Audio model serving that caches a processor for every distinct request-coCVE-2026-114052 · Other software · vllm-project · reported by VulnCheckHigh 8.7
vLLM after 0.24.0 and before 0.31.0 contains a memory exhaustion vulnerability in MOSS-Audio model serving that caches a processor for every distinct request-controlled mm_processor_kwargs value in an unbounded dictionary. Unauthenticated attackers can send distinct or oversized mel_config values to the /tokenize endpoint, which bypasses --api-key, to exhaust host memory and crash the API process.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD recordgithub.comvulncheck.com
Oct 20, 2026vLLM before 0.31.0 skips the min_tokens <= max_tokens check for /inference/v1/generate requests with a null max_tokens, so the EngineCore input thread later dieCVE-2026-114051 · Other software · vllm-project · reported by VulnCheckHigh 7.1
vLLM before 0.31.0 skips the min_tokens <= max_tokens check for /inference/v1/generate requests with a null max_tokens, so the EngineCore input thread later dies on an uncaught VLLMValidationError while /health still returns 200. Unauthenticated remote attackers can send one request with an oversized min_tokens to hang all subsequent generation requests until manual restart.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD recordgithub.comvulncheck.com
Oct 20, 2026vLLM before 0.31.0 uses a client-supplied multimodal uuid verbatim as cache identity with no consistency check on cache hits, allowing engine crashes.CVE-2026-114050 · Other software · vllm-project · reported by VulnCheckHigh 7.1
vLLM before 0.31.0 uses a client-supplied multimodal uuid verbatim as cache identity with no consistency check on cache hits, allowing engine crashes. Unauthenticated attackers can desynchronize the P0/P1 processor caches to trigger a fatal token-placeholder mismatch, killing the entire serving instance until restarted.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD recordgithub.comvulncheck.com
Oct 20, 2026vLLM before 0.31.0 contains a cross-request mask leak in InputBatch.condense() that allows remote attackers to constrain other clients' outputs by leaving staleCVE-2026-114049 · Other software · vllm-project · reported by VulnCheckMedium 6.3
vLLM before 0.31.0 contains a cross-request mask leak in InputBatch.condense() that allows remote attackers to constrain other clients' outputs by leaving stale allowed_token_ids rows. Attackers can submit concurrent constrained generate requests, including via unauthenticated /invocations, so recycled batch slots apply their token allowlist to victims' unrestricted requests.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD recordgithub.comvulncheck.com
Oct 12, 2026vLLM before 0.29.0 passes uploaded audio to the PyAV/FFmpeg av.open() fallback with only a file-size check, exposing FFmpeg's IAMF parser heap overflow.CVE-2026-109311 · Other software · vllm-project · reported by VulnCheckHigh 7.1
vLLM before 0.29.0 passes uploaded audio to the PyAV/FFmpeg av.open() fallback with only a file-size check, exposing FFmpeg's IAMF parser heap overflow. Authenticated remote attackers can repeatedly upload malformed IAMF containers to /v1/audio/transcriptions, corrupting the heap and crashing the API process.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD recordgithub.comvulncheck.com
Oct 6, 2026A security flaw has been discovered in vllm-project vLLM up to 0.31.0.CVE-2026-105922 · Other software · vllm-project · reported by VulDBLow 2.1
A security flaw has been discovered in vllm-project vLLM up to 0.31.0. This impacts the function get_token_bin_counts_and_mask of the file vllm/model_executor/layers/utils.py of the component Penalty Handler. Performing a manipulation results in denial of service. Remote exploitation of the attack is possible. The exploit has been released to the public and may be used for attacks. The project was informed of the problem early through an issue report but has not responded yet.
- Product
- vLLM
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD recordgist.github.comgithub.com
Oct 6, 2026A security vulnerability has been detected in vllm-project vLLM up to 0.31.0.CVE-2026-105775 · Other software · vllm-project · reported by VulDBLow 2.1
A security vulnerability has been detected in vllm-project vLLM up to 0.31.0. This impacts the function conv_ssm_forward of the file vllm/model_executor/layers/mamba/mamba_mixer2.py of the component Completions Request Handler. The manipulation leads to out-of-bounds read. The attack is possible to be carried out remotely. The exploit has been disclosed publicly and may be used. The project was informed of the problem early through an issue report but has not responded yet.
- Product
- vLLM
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Oct 5, 2026vLLM is an inference and serving engine for large language models.CVE-2026-105760 · Other software · vllm-project · reported by GitHubMedium 5.3
vLLM is an inference and serving engine for large language models. Prior to 0.30.0, a caller can use the request-level media_io_kwargs field to select the GLMGA video backend and supply large values for the fps and max_frames options without a strict work ceiling. GLMGA constructs and deduplicates an attacker-sized pre-decode frame-index list, allowing a compact request and tiny valid video to consume disproportionate CPU time and memory in the shared media-loading executor. This issue is fixed in version 0.30.0.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Oct 5, 2026vLLM is an inference and serving engine for large language models.CVE-2026-105759 · Other software · vllm-project · reported by GitHubMedium 5.9
vLLM is an inference and serving engine for large language models. Prior to 0.30.0, the Rust frontend's track_http_metrics middleware records the raw HTTP method token as a Prometheus label for requests reaching registered routes. An unauthenticated attacker can send unique arbitrary method tokens to unguarded routes such as /tokenize, causing Prometheus's Family::get_or_create function to permanently create counter and histogram label sets. Those label sets increase process memory usage and enlarge the /metrics response until the service or monitoring path is exhausted. This issue is fixed in version 0.30.0.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Oct 5, 2026vLLM is an inference and serving engine for large language models.CVE-2026-105758 · Other software · vllm-project · reported by GitHubMedium 5.3
vLLM is an inference and serving engine for large language models. From 0.24.0 until 0.30.0, the Qwen2VLVideoBackend and Qwen3VLVideoBackend classes accept request-level values for the media_io_kwargs.video.max_frames and media_io_kwargs.video.fps fields without enforcing server-side ceilings. An unauthenticated caller can submit these values to the /tokenize endpoint, causing the sampler to decode every frame selected from attacker-controlled video input, consume disproportionate frontend memory, and potentially terminate the API process before scheduling or admission control. The Rust frontend is not affected because it rejects the media_io_kwargs field. This issue is fixed in version 0.30.0.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Oct 5, 2026vLLM is an inference and serving engine for large language models.CVE-2026-105757 · Other software · vllm-project · reported by GitHubMedium 6.5
vLLM is an inference and serving engine for large language models. Prior to 0.30.0, structured-output request failures can escape request-scoped validation and reach the EngineCore fatal-error path. A per-request backend mismatch can re-raise a grammar compilation exception, padding produced by the ngram_gpu speculative-decoding mode can pass a negative token to guidance validation, and the Rust frontend can admit empty structured-output values that the Python frontend rejects, allowing ordinary constrained-generation requests to terminate the shared engine. This issue is fixed in version 0.30.0.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Oct 5, 2026vLLM is an inference and serving engine for large language models.CVE-2026-105756 · Other software · vllm-project · reported by GitHubMedium 6.5
vLLM is an inference and serving engine for large language models. Prior to 0.30.0, OpenAI-compatible request models accept a non-empty cache_salt value without enforcing the character and length restrictions required by the IPCCacheServerKey consumer in LMCache-MP. On deployments using the LMCache-MP connector, a salt that contains a forbidden character or exceeds the permitted length can raise an uncaught ValueError during scheduler cache lookup, causing EngineCore to terminate and denying service to all concurrent users. This issue is fixed in version 0.30.0.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Oct 5, 2026vLLM is an inference and serving engine for large language models.CVE-2026-105755 · Other software · vllm-project · reported by GitHubMedium 4.2
vLLM is an inference and serving engine for large language models. Prior to 0.30.0, flash late-interaction scoring at the /score and /rerank endpoints derives each worker's query_key value from the caller-controlled X-Request-Id header. A concurrent request that reuses a victim's identifier can overwrite the cached query embedding so the victim's documents are scored against the attacker's query, and shared use counters can also cause a late-interaction cache-miss error. This issue is fixed in version 0.30.0.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Oct 5, 2026vLLM is an inference and serving engine for large language models.CVE-2026-105754 · Other software · vllm-project · reported by GitHubMedium 6.5
vLLM is an inference and serving engine for large language models. Prior to 0.30.0, the /inference/v1/generate endpoint in the disaggregated scale-out path accepts caller-supplied tensors in the features.kwargs_data field, cache identifiers in the features.mm_hashes field, ranges in the features.mm_placeholders field, and wire-selected multimodal field processors without rebinding them to the active model renderer contract. Forged grid geometry, field types, or non-positive placeholder lengths can terminate the shared EngineCore; when an attacker knows or can induce a victim's content hash, forged cache hashes can poison or retrieve cross-request encoder-cache state; and dropped sparse placeholder masks can alter replayed transport semantics. This issue is fixed in version 0.30.0.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Oct 5, 2026vLLM is an inference and serving engine for large language models.CVE-2026-105753 · Other software · vllm-project · reported by GitHubMedium 6.5
vLLM is an inference and serving engine for large language models. Prior to 0.28.0, the default mirrored multimodal LRU cache can commit a media hash in the frontend sender cache during multimodal rendering and before engine admission, while the engine receiver cache never receives the payload if that request is rejected. A later request reusing the same media hash causes MultiModalProcessorSenderCache to send no payload and MultiModalReceiverCache to reach an assertion with the message "Expected a cached item," producing a shared-service availability failure. This issue is fixed in version 0.28.0.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Oct 5, 2026vLLM is an inference and serving engine for large language models.CVE-2026-105752 · Other software · vllm-project · reported by GitHubLow 3.1
vLLM is an inference and serving engine for large language models. Prior to 0.30.0, Harmony tool continuations submitted through "POST /v1/responses" requests rebuild the next-turn engine input without preserving the cache_salt value, placing the continuation prefix in the global unsalted cache namespace even when the caller enabled salting. On deployments with prefix caching enabled, which is the default, an authenticated tenant who can reconstruct a victim's low-entropy post-tool history can submit the same continuation and use the cached_tokens_per_turn count to determine whether the prefix was previously processed, defeating the intended tenant isolation of salted prefix caching. This issue is fixed in version 0.30.0.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Sep 30, 2026A flaw has been found in vllm-project vLLM up to 0.26.0.CVE-2026-103241 · Other software · vllm-project · reported by VulDBMedium 5.5
A flaw has been found in vllm-project vLLM up to 0.26.0. This vulnerability affects unknown code of the file rust/src/parser/src/unified/gemma4.rs of the component Gemma4UnifiedParser. Executing a manipulation can lead to denial of service. The attack may be launched remotely. The exploit has been published and may be used. Upgrading to version 0.29.1rc0 is able to resolve this issue. This patch is called 3439bad37e68ba9755a46f4f6b44a4aeaf1f60a9. Upgrading the affected component is advised.
- Product
- vLLM
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD recordgist.github.comgithub.com
Sep 26, 2026vLLM before 0.29.0 accepts user-controlled stop_token_ids on the OpenAI-compatible POST /v1/completions and POST /v1/chat/completions endpoints but validates onCVE-2026-100654 · Other software · vllm-project · reported by VulnCheckHigh 7.1
vLLM before 0.29.0 accepts user-controlled stop_token_ids on the OpenAI-compatible POST /v1/completions and POST /v1/chat/completions endpoints but validates only that the values are integers, not that each token id is within the model vocabulary/logits range. When min_tokens > 0, the stop token ids are used as logits indices to suppress stop tokens, so an out-of-range id reaches a CUDA indexing operation (index_put_) and triggers a device-side assertion. An authenticated API user can send a single malformed completion request that returns 500 Internal Server Error and puts EngineCore into a fatal state, causing subsequent requests to fail until the service is restarted (denial of service).
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD recordgithub.comvulncheck.com
Sep 26, 2026vLLM is an inference and serving engine for large language models.CVE-2026-100653 · Other software · vllm-project · reported by VulnCheckHigh 8.3
vLLM is an inference and serving engine for large language models. In versions from 0.22.1 through 0.28.0, the operator-supplied model revision pin (--revision / --code-revision) is not propagated to several Hugging Face artifact loads for the FunAudioChat and Tarsier2 architectures: the WhisperFeatureExtractor and speech_tokenizer PreTrainedTokenizerFast loads in vllm/model_executor/models/funaudiochat.py and the Qwen2VLConfig.from_pretrained call used by Tarsier2ProcessingInfo in vllm/model_executor/models/qwen2_vl.py. As a result, deployments pinned to a reviewed revision still resolve these behavior-affecting processor, tokenizer, and config artifacts from the repository's default revision, so a later change to the upstream default branch can alter audio preprocessing, speech tokenizer behavior, or Tarsier2 configuration without any change to the operator's configured pin. This is a supply-chain integrity and reproducibility failure for pinned deployments; it is residual to the earlier fix tracked as GHSA-3ww4-5jv9-j5gm / CVE-2026-47155 and does not constitute remote code execution or a trust_remote_code=False bypass. The issue is fixed in version 0.28.0.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Sep 26, 2026vLLM versions 0.22.0 through 0.23.0 fail to validate stop_token_ids against vocabulary bounds in Rust HTTP and gRPC frontends, allowing out-of-vocabulary token CVE-2026-100652 · Other software · vllm-project · reported by VulnCheckHigh 8.2
vLLM versions 0.22.0 through 0.23.0 fail to validate stop_token_ids against vocabulary bounds in Rust HTTP and gRPC frontends, allowing out-of-vocabulary token IDs to reach MinTokensLogitsProcessor. Attackers can submit requests with min_tokens greater than zero and out-of-vocabulary stop_token_ids to trigger CUDA tensor indexing failures that leave EngineCore in a fatal state requiring service restart.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD recordgithub.comvulncheck.com
Sep 26, 2026vLLM before 0.29.0 fails to enforce decoder prompt-length validation on the disaggregated serving endpoint /inference/v1/generate.CVE-2026-100651 · Other software · vllm-project · reported by VulnCheckHigh 7.1
vLLM before 0.29.0 fails to enforce decoder prompt-length validation on the disaggregated serving endpoint /inference/v1/generate. When the request contains a 'features' (multimodal) payload, vllm/entrypoints/serve/disagg/serving.py builds a multimodal EngineInput directly from the caller-supplied token_ids, and GenerateRequest.token_ids (vllm/entrypoints/serve/disagg/protocol.py) is not checked against model_config.max_model_len. For multimodal processors that report skip_prompt_length_check=True (for example Nemotron Parse, Whisper, and FireRedLID), InputProcessor._validate_prompt_len() returns immediately for both encoder and decoder prompts, so an overlong prompt becomes an EngineCoreRequest and reaches the worker input-batch copy into a fixed max_model_len-wide NumPy row. A client able to reach the endpoint on an affected model configuration can therefore submit an overlong token_ids list to trigger a worker failure and denial of service. Fixed in 0.29.0.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD recordgithub.comvulncheck.com
Sep 26, 2026vLLM through 0.29.0 fetches and fully materializes remote or inline media before enforcing its documented media controls (the VLLM_MAX_AUDIO_CLIP_FILESIZE_MB coCVE-2026-100650 · Other software · vllm-project · reported by VulnCheckHigh 7.1
vLLM through 0.29.0 fetches and fully materializes remote or inline media before enforcing its documented media controls (the VLLM_MAX_AUDIO_CLIP_FILESIZE_MB compressed-audio size cap, default 25 MB, and the per-modality --limit-mm-per-prompt item limits). Across four ingress paths — the shared media-acquisition layer (HTTPConnection.get_bytes()/async_get_bytes()), the chat completions audio_url/base64 path, the batch speech runner, and the Rust frontend POST /tokenize route — the server reads the entire HTTP response body, base64-decodes the inline payload, or spawns one fetch/decode task per media part, and only then applies the limit (or, on some paths, never applies it). A remote attacker can therefore cause the API server or batch-runner process to allocate memory and consume outbound bandwidth proportional to an attacker-chosen body size or media item count before the request is rejected, resulting in pre-inference memory and bandwidth exhaustion (denial of service). The chat and batch surfaces require an API key when one is configured; the Rust frontend /tokenize route is unauthenticated by design. There is no code execution or data disclosure impact.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Sep 26, 2026vLLM before 0.29.0 contains a resource-limit bypass vulnerability in PyNvVideoCodec decoder allocation where sampler subclass shadowing allows independent countCVE-2026-100649 · Other software · vllm-project · reported by VulnCheckMedium 6.3
vLLM before 0.29.0 contains a resource-limit bypass vulnerability in PyNvVideoCodec decoder allocation where sampler subclass shadowing allows independent counter increments. Unauthenticated attackers can select different sampler subclasses in video requests to exceed configured decoder limits and exhaust unaccounted GPU memory.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD recordgithub.comvulncheck.com
Sep 26, 2026vllm before 0.29.0 fails to enforce VLLM_MAX_AUDIO_CLIP_FILESIZE_MB limit in multimodal chat audio decoding, allowing unauthenticated clients to bypass file sizCVE-2026-100648 · Other software · vllm-project · reported by VulnCheckMedium 6.9
vllm before 0.29.0 fails to enforce VLLM_MAX_AUDIO_CLIP_FILESIZE_MB limit in multimodal chat audio decoding, allowing unauthenticated clients to bypass file size restrictions. Attackers can submit oversized audio files through chat endpoints to consume excessive memory and CPU resources during decoding.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD recordgithub.comvulncheck.com
Sep 26, 2026vLLM versions before 0.29.0 contain a denial-of-service vulnerability in the cache_salt parameter accepted on OpenAI-compatible and Anthropic API endpoints, whiCVE-2026-100647 · Other software · vllm-project · reported by VulnCheckMedium 6.9
vLLM versions before 0.29.0 contain a denial-of-service vulnerability in the cache_salt parameter accepted on OpenAI-compatible and Anthropic API endpoints, which lacks maximum length validation and is processed on the single EngineCore scheduler thread. Unauthenticated attackers can send HTTP requests with multi-hundred-megabyte salt values that trigger expensive pickle serialization and SHA-256 hashing, stalling the scheduler thread and denying service to all concurrent requests.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD recordgithub.comvulncheck.com
Sep 21, 2026vLLM Mooncake connector through 0.29.0 fails to properly manage GPU KV cache block ownership when concurrent child requests share a single transfer ID in prefilCVE-2026-94627 · Other software · vllm-project · reported by VulnCheckHigh 8.7
vLLM Mooncake connector through 0.29.0 fails to properly manage GPU KV cache block ownership when concurrent child requests share a single transfer ID in prefill/decode disaggregated deployments. Attackers can trigger GPU memory exhaustion by submitting completion requests with multiple prompts, causing orphaned KV cache blocks to accumulate until process restart and eventually preventing legitimate requests from executing.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Sep 21, 2026vLLM through 0.29.0 fails to validate the tp_size parameter in kv_transfer_params on OpenAI-compatible completion endpoints, allowing attackers to allocate unboCVE-2026-94626 · Other software · vllm-project · reported by VulnCheckHigh 8.7
vLLM through 0.29.0 fails to validate the tp_size parameter in kv_transfer_params on OpenAI-compatible completion endpoints, allowing attackers to allocate unbounded memory. Attackers can supply arbitrary tp_size values in prefill/decode disaggregated deployments to exhaust memory and trigger kernel OOM-kill of the decode worker process.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Sep 21, 2026vLLM through 0.29.0 contains a resource exhaustion vulnerability in MooncakeConnector where rejected prefill requests create ownerless transfer placeholders thaCVE-2026-94625 · Other software · vllm-project · reported by VulnCheckMedium 6.9
vLLM through 0.29.0 contains a resource exhaustion vulnerability in MooncakeConnector where rejected prefill requests create ownerless transfer placeholders that are never reclaimed. Attackers can send rejected requests to exhaust sender task pools, causing valid requests to be delayed by up to 480 seconds while health checks continue returning success.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Sep 21, 2026vLLM through 0.29.0 contains a denial of service vulnerability in P2P KV offloading when OffloadingConnector is configured with TieringOffloadingSpec and a peerCVE-2026-94624 · Other software · vllm-project · reported by VulnCheckHigh 8.7
vLLM through 0.29.0 contains a denial of service vulnerability in P2P KV offloading when OffloadingConnector is configured with TieringOffloadingSpec and a peer-to-peer secondary tier. Attackers can supply arbitrary remote host and port values in kv_transfer_params to create unreachable peer sessions that retain ZeroMQ sockets until the context quota is exhausted, causing an uncaught ZMQError that crashes EngineCore and stops all inference.
- Product
- vllm
- What to do
- Apply the vendor’s update. The NVD record links to the advisory.
Where this comes from: the exploited list is CISA’s Known Exploited Vulnerabilities catalog. Everything else is the whole National Vulnerability Database, where each record is published by the team that found the bug (Wordfence, Patchstack, WPScan, the vendors themselves and others); we show who. Exploit odds are FIRST’s EPSS, updated daily. Titles and categories are ours, so check the linked record before acting on a detail. For a write-up of the big ones, see our advisories. Every vendor, and every WordPress plugin and theme with a recorded flaw, also has its own page. All of it is free to reuse as JSON, CSV or RSS.
Weekly security brief
The flaws that hit local websites, every Friday.
One short email: what got exploited this week, which WordPress plugins broke, and what to check. Unsubscribe in one click.
Tell us what you're building.
A website, your marketing, a security check, an app, or the thing you've been running out of a spreadsheet for three years. A few lines is plenty.
Not ready to talk yet? Run the free website check and see what Google sees.