- 6.5MediumCVSS base score
- 0.58%Exploitation in 30 daysEPSS · top 52% of all CVEs
- NoKnown exploitednot in CISA's KEV catalog
- YesPublic exploitsa reference or SSVC says so
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.
Public exploit
An exploit or proof of concept is public: anyone can try it against an unpatched system. A reference of the record is tagged as an exploit. CISA's SSVC rates its exploitation: a proof of concept is public.
Show the exploits1
References tagged exploit1 reference
- https://github.com/vllm-project/vllm/security/advisories/GHSA-8pw2-6jv3-mj5j (opens in a new tab)x_refsource_CONFIRMexploit
Timeline
ID reserved
By GitHub_M
- 44 days later
· 44 days later
Published
Changes after publication (a new score, a public exploit, EPSS crossing 1, 10 or 50 %, leaving KEV) are kept since November 2026; earlier ones are not known.
Scores
| Scored by | CVSS | Score | Vector |
|---|---|---|---|
| GitHub_MCNA (assigner) | 3.1 | Medium6.5 | CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H |
| NVDNVD · Analyzed | 3.1 | Medium6.5 | CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H |
- EPSS (exploitation in 30 days)
- 0.58% · top 52% of all CVEs (11 Nov 2026)
- CISA SSVC
- Exploitation: poc · Automatable: no · Technical impact: partial
The list uses the assigner's score when there is one, then CISA's, then NVD's: Medium 6.5.
Affected products
- affected < 0.28.0
Affected packages
From reviewed advisories of the GitHub Advisory Database.
References
References4
- https://github.com/vllm-project/vllm/security/advisories/GHSA-8pw2-6jv3-mj5j (opens in a new tab)x_refsource_CONFIRMexploit
- https://github.com/vllm-project/vllm/pull/47259 (opens in a new tab)x_refsource_MISC
- https://github.com/vllm-project/vllm/commit/283893c72292ede38d277e3cd2b9b64c3e4f1dda (opens in a new tab)x_refsource_MISC
- https://github.com/vllm-project/vllm/commit/ba22152096b2484faa3579624a253d54804d876d (opens in a new tab)x_refsource_MISC
Sources
The same CVE in the primary databases and aggregators, to check against.
- CVE.org record (opens in a new tab)
- CVE List JSON (GitHub) (opens in a new tab)
- NVD (opens in a new tab)
- FIRST EPSS (opens in a new tab)
- EUVD EUVD-2026-80935 (opens in a new tab)
- OSV.dev (opens in a new tab)
- GitHub Advisory GHSA-8pw2-6jv3-mj5j (opens in a new tab)
- CIRCL Vulnerability-Lookup (opens in a new tab)
Badge for a README or a ticket
[](https://cveledger.app/cve/CVE-2026-69147)Record last updated .