{"id":"CVE-2026-100651","url":"https://www.cyber-defence.io/tools/cve/CVE-2026-100651","generatedAt":"2026-10-28T05:28:23.054Z","title":"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.","description":"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.","published":"2026-09-26T14:16:48.000Z","lastModified":"2026-10-06T19:30:26.000Z","status":"Analyzed","sourceIdentifier":"disclosure@vulncheck.com","cvss":{"version":"4.0","score":7.1,"severity":"HIGH","vector":"CVSS:4.0/AV:N/AC:L/AT:N/PR:L/UI:N/VC:N/VI:N/VA:H/SC:N/SI:N/SA:N/E:X/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/M"},"cwe":["CWE-400"],"affected":["vllm/vllm"],"epss":{"score":0.00384,"percentile":0.3231,"date":"2026-10-27","history":[{"date":"2026-09-27","score":0.00307,"percentile":0.21045}]},"kev":{"listed":false},"exploits":{"count":0,"verified":false,"firstPublished":null,"source":"Exploit-DB (https://gitlab.com/exploit-database/exploitdb)","entries":[]},"verdict":{"level":"high","text":"High impact if exploited, but EPSS currently rates exploitation as unlikely (0.38%). Schedule it in the normal patch cycle and watch for a rise in EPSS or a public exploit."},"changes":[{"kind":"new","label":"New CVE","at":"2026-09-26T14:16:48.000Z","detail":{"score":"7.1","severity":"HIGH","publishedAt":"2026-09-26T14:16:47.663Z"},"summary":"New CVE, high severity (CVSS 7.1)."}],"references":[{"url":"https://github.com/vllm-project/vllm/security/advisories/GHSA-3mqx-f33v-vgp9","tags":["Exploit","Vendor Advisory"],"source":"disclosure@vulncheck.com"},{"url":"https://www.vulncheck.com/advisories/vllm-before-0.29.0-denial-of-service-via-decoder-prompt-length-bypass","tags":["Third Party Advisory"],"source":"disclosure@vulncheck.com"},{"url":"https://github.com/vllm-project/vllm/security/advisories/GHSA-3mqx-f33v-vgp9","tags":["Exploit","Vendor Advisory"],"source":"134c704f-9b21-4f2e-91b3-4a467353bcc0"}],"sources":{"nvd":"https://nvd.nist.gov/vuln/detail/CVE-2026-100651","epss":"https://www.first.org/epss/","kev":"https://www.cisa.gov/known-exploited-vulnerabilities-catalog","exploitdb":"https://www.exploit-db.com/"},"licence":"CC BY 4.0 — link back to the CVE Explorer if you publish the results; upstream data remains subject to its own terms."}