{"id":"CVE-2026-69147","state":"PUBLISHED","title":"vLLM: Request-selected PyNvVideoCodec GPU decode bypasses static VRAM reservation","description":"vLLM is an inference and serving engine for large language models. Prior to 0.28.0, request bodies for Chat Completions and Responses can set media_io_kwargs.video.video_backend to pynvvideocodec, and MediaConnector.fetch_video forwards that choice to VideoMediaIO even when startup configuration selected a software decoder. The engine's _reserve_mm_ipc_gpu_memory logic budgets decoder memory only from static configuration, so the request-selected VIDEO_LOADER_REGISTRY backend can create a CUDA context, decoder surfaces, and decoded-frame allocations that were not removed from the engine's KV-cache budget. An attacker able to submit video requests to a video-capable GPU deployment with PyNvVideoCodec installed can exhaust shared GPU memory, causing request failures, worker crashes, or denial of service. The first release containing the fix is version 0.28.0.","publishedAt":"2026-09-16T17:49:20.427Z","updatedAt":"2026-09-16T18:37:14.118Z","severity":{"score":6.5,"label":"Medium","version":"3.1","vector":"CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H","source":"GitHub_M"},"assigner":"GitHub_M","products":["vllm-project vllm"],"affected":[{"vendor":"vllm-project","product":"vllm","versions":[{"version":"< 0.28.0","status":"affected"}]}],"references":["https://github.com/vllm-project/vllm/security/advisories/GHSA-8pw2-6jv3-mj5j","https://github.com/vllm-project/vllm/pull/47259","https://github.com/vllm-project/vllm/commit/283893c72292ede38d277e3cd2b9b64c3e4f1dda","https://github.com/vllm-project/vllm/commit/ba22152096b2484faa3579624a253d54804d876d"],"weaknesses":["CWE-400","CWE-770"],"ingestedAt":"2026-10-29T20:12:49.727Z","source":{"repository":"CVEProject/cvelistV5","path":"cves/2026/69xxx/CVE-2026-69147.json","commit":"556c9d3a8cd17e218bb8e452ba7cd619a384ff47"},"raw":{"dataType":"CVE_RECORD","dataVersion":"5.2","cveMetadata":{"cveId":"CVE-2026-69147","state":"PUBLISHED","datePublished":"2026-09-16T17:49:20.427Z","dateUpdated":"2026-09-16T18:37:14.118Z","assignerShortName":"GitHub_M","assignerOrgId":"a0819718-46f1-4df5-94e2-005712e83aaa","dateReserved":"2026-08-03T15:20:30.218Z"},"containers":{"cna":{"title":"vLLM: Request-selected PyNvVideoCodec GPU decode bypasses static VRAM reservation","descriptions":[{"lang":"en","value":"vLLM is an inference and serving engine for large language models. Prior to 0.28.0, request bodies for Chat Completions and Responses can set media_io_kwargs.video.video_backend to pynvvideocodec, and MediaConnector.fetch_video forwards that choice to VideoMediaIO even when startup configuration selected a software decoder. The engine's _reserve_mm_ipc_gpu_memory logic budgets decoder memory only from static configuration, so the request-selected VIDEO_LOADER_REGISTRY backend can create a CUDA context, decoder surfaces, and decoded-frame allocations that were not removed from the engine's KV-cache budget. An attacker able to submit video requests to a video-capable GPU deployment with PyNvVideoCodec installed can exhaust shared GPU memory, causing request failures, worker crashes, or denial of service. The first release containing the fix is version 0.28.0."}],"affected":[{"vendor":"vllm-project","product":"vllm","versions":[{"version":"< 0.28.0","status":"affected"}]}],"references":[{"url":"https://github.com/vllm-project/vllm/security/advisories/GHSA-8pw2-6jv3-mj5j","name":"https://github.com/vllm-project/vllm/security/advisories/GHSA-8pw2-6jv3-mj5j","tags":["x_refsource_CONFIRM"]},{"url":"https://github.com/vllm-project/vllm/pull/47259","name":"https://github.com/vllm-project/vllm/pull/47259","tags":["x_refsource_MISC"]},{"url":"https://github.com/vllm-project/vllm/commit/283893c72292ede38d277e3cd2b9b64c3e4f1dda","name":"https://github.com/vllm-project/vllm/commit/283893c72292ede38d277e3cd2b9b64c3e4f1dda","tags":["x_refsource_MISC"]},{"url":"https://github.com/vllm-project/vllm/commit/ba22152096b2484faa3579624a253d54804d876d","name":"https://github.com/vllm-project/vllm/commit/ba22152096b2484faa3579624a253d54804d876d","tags":["x_refsource_MISC"]}],"metrics":[{"cvssV3_1":{"attackComplexity":"LOW","attackVector":"NETWORK","availabilityImpact":"HIGH","baseScore":6.5,"baseSeverity":"MEDIUM","confidentialityImpact":"NONE","integrityImpact":"NONE","privilegesRequired":"LOW","scope":"UNCHANGED","userInteraction":"NONE","vectorString":"CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H","version":"3.1"}}],"problemTypes":[{"descriptions":[{"cweId":"CWE-400","lang":"en","description":"CWE-400: Uncontrolled Resource Consumption","type":"CWE"}]},{"descriptions":[{"cweId":"CWE-770","lang":"en","description":"CWE-770: Allocation of Resources Without Limits or Throttling","type":"CWE"}]}],"providerMetadata":{"shortName":"GitHub_M","orgId":"a0819718-46f1-4df5-94e2-005712e83aaa","dateUpdated":"2026-09-16T17:49:20.427Z"},"source":{"advisory":"GHSA-8pw2-6jv3-mj5j","discovery":"UNKNOWN"}},"adp":[{"title":"CISA ADP Vulnrichment","references":[{"url":"https://github.com/vllm-project/vllm/security/advisories/GHSA-8pw2-6jv3-mj5j","tags":["exploit"]}],"metrics":[{"other":{"type":"ssvc","content":{"timestamp":"2026-09-16T18:37:08.090589Z","id":"CVE-2026-69147","options":[{"Exploitation":"poc"},{"Automatable":"no"},{"Technical Impact":"partial"}],"role":"CISA Coordinator","version":"2.0.3"}}}],"providerMetadata":{"shortName":"CISA-ADP","orgId":"134c704f-9b21-4f2e-91b3-4a467353bcc0","dateUpdated":"2026-09-16T18:37:14.118Z"}}]}}}