PT-2026-120179

Name of the Vulnerable Software and Affected Versions vLLM versions 0.7.3 through 0.30.x Description When trust remote code is enabled, the software forwards the request-supplied mm processor kwargs.code revision to the Transformers AutoProcessor.from pretrained function. This allows clients to...

PT-2026-120178

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...

vLLM code injection vulnerability

Versions vLLM 0.7.3 through 0.31.0 contain a code injection vulnerability. This vulnerability arises when the mmprocessorkwargs.coderevision parameter provided by the request is passed to Transformers AutoProcessor.frompretrained. When trustremotecode is enabled, it allows clients to specify whic...

vLLM: Qwen2-VL / Qwen3-VL video samplers bound on request-controlled max_frames, which the num_frames ceiling does not reach

SummaryAn unauthenticated remote attacker can exhaust the memory of the vLLM API-server process by raising the request-level mediaiokwargs.video.maxframes and fps knobs on any deployment serving a Qwen2-VL or Qwen3-VL model. 74 extra bytes of JSON took the server's peak RSS from 2 271 MiB to 13 6...

EUVD-2026-92848

vLLM: Qwen2-VL / Qwen3-VL video samplers bound on request-controlled maxframes, which the numframes ceiling does not reach...

vLLM: Qwen2-VL / Qwen3-VL video samplers bound on request-controlled max_frames, which the num_frames ceiling does not reach

Summary An unauthenticated remote attacker can exhaust the memory of the vLLM API-server process by raising the request-level mediaiokwargs.video.maxframes and fps knobs on any deployment serving a Qwen2-VL or Qwen3-VL model. 74 extra bytes of JSON took the server's peak RSS from 2 271 MiB to 13...

CVE-2026-105758

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 mediaiokwargs.video.maxframes and mediaiokwargs.video.fps fields without enforcing server-side ceilings. An...

EUVD-2026-92850

vLLM is an inference and serving engine for large language models. Prior to 0.30.0, the Rust frontend's trackhttpmetrics 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 ...

CVE-2026-105759

vLLM is an inference and serving engine for large language models. Prior to 0.30.0, the Rust frontend's trackhttpmetrics 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 ...

CVE-2026-105759

The vLLM inference and serving engine for large language models is vulnerable to a Denial of Service (DoS) attack in versions prior to 0.30.0 . The root cause lies in the Rust frontend's track_http_metrics middleware, which records raw HTTP method tokens as Prometheus labels without validation or...

CVE-2026-105758

The vLLM inference engine (versions 0.24.0 through 0.30.0 ) is vulnerable to a resource exhaustion issue within the Qwen2VLVideoBackend and Qwen3VLVideoBackend classes. The software fails to enforce server-side ceilings on media_io_kwargs.video.max_frames and media_io_kwargs.video.fps fields prov...

CVE-2026-105758 vLLM: Qwen2-VL / Qwen3-VL video samplers bound on request-controlled max_frames, which the num_frames ceiling does not reach

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 mediaiokwargs.video.maxframes and mediaiokwargs.video.fps fields without enforcing server-side ceilings. An...

CVE-2026-105758 vLLM: Qwen2-VL / Qwen3-VL video samplers bound on request-controlled max_frames, which the num_frames ceiling does not reach

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 mediaiokwargs.video.maxframes and mediaiokwargs.video.fps fields without enforcing server-side ceilings. An...

CVE-2026-105758 vLLM: Qwen2-VL / Qwen3-VL video samplers bound on request-controlled max_frames, which the num_frames ceiling does not reach

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 mediaiokwargs.video.maxframes and mediaiokwargs.video.fps fields without enforcing server-side ceilings. An...

CVE-2026-105758

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 mediaiokwargs.video.maxframes and mediaiokwargs.video.fps fields without enforcing server-side ceilings. An...

CVE-2026-105758: Allocation of Resources Without Limits or Throttling

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 mediaiokwargs.video.maxframes and mediaiokwargs.video.fps fields without enforcing server-side ceilings. An...

PT-2026-106270

Name of the Vulnerable Software and Affected Versions vLLM versions 0.24.0 through 0.29.1rc0 Description An unauthenticated remote attacker can cause a denial of service by exhausting the memory of the API-server process. This occurs when deployments serving Qwen2-VL or Qwen3-VL models accept...

PYSEC-2026-4184

vLLM through 0.29.0 fetches and fully materializes remote or inline media before enforcing its documented media controls the VLLMMAXAUDIOCLIPFILESIZEMB compressed-audio size cap, default 25 MB, and the per-modality --limit-mm-per-prompt item limits. Across four ingress paths — the shared...

PYSEC-2026-4184

vLLM through 0.29.0 fetches and fully materializes remote or inline media before enforcing its documented media controls the VLLMMAXAUDIOCLIPFILESIZEMB compressed-audio size cap, default 25 MB, and the per-modality --limit-mm-per-prompt item limits. Across four ingress paths — the shared...

CVE-2026-100650

vLLM through 0.29.0 fetches and fully materializes remote or inline media before enforcing its documented media controls the VLLMMAXAUDIOCLIPFILESIZEMB compressed-audio size cap, default 25 MB, and the per-modality --limit-mm-per-prompt item limits. Across four ingress paths — the shared...