{"id":"CVE-2026-54234","url":"https://www.cyber-defence.io/tools/cve/CVE-2026-54234","generatedAt":"2026-10-28T00:25:36.561Z","title":"Prior to 0.24.0, a frontend-legal multi-request speculative decoding workload can cause the rejection sampler to produce a recovered token equal to the model vocabulary size boundary value, which is then converted to negative one when the engine selects…","description":"vLLM is a high-throughput and memory-efficient inference and serving engine for LLMs. Prior to 0.24.0, a frontend-legal multi-request speculative decoding workload can cause the rejection sampler to produce a recovered token equal to the model vocabulary size boundary value, which is then converted to negative one when the engine selects the next live token for a request and is written back into the drafter's input ids; that out-of-vocabulary value is later consumed by the model's embedding and attention path and crashes the engine worker with a GPU device-side assertion. The same triggering request sequence is reachable through the public gRPC Generate and Abort endpoints, so a remote client that can send generation requests can crash the shared engine worker, aborting concurrent requests and causing a service-wide denial of service for other clients of the deployment until the worker is restarted. This issue is fixed in version 0.24.0.","published":"2026-07-06T21:16:56.000Z","lastModified":"2026-10-25T20:10:00.000Z","status":"Analyzed","sourceIdentifier":"security-advisories@github.com","cvss":{"version":"3.1","score":7.5,"severity":"HIGH","vector":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H"},"cwe":["CWE-20","CWE-1284"],"affected":["vllm/vllm"],"epss":{"score":0.00616,"percentile":0.4924,"date":"2026-10-26","history":[]},"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 (0.62%). Schedule it in the normal patch cycle and watch for a rise in EPSS or a public exploit."},"changes":[],"references":[{"url":"https://github.com/vllm-project/vllm/commit/8a5cf1ccd65e8ac7635c402c1ec0b08988bc26ca","tags":["Patch"],"source":"security-advisories@github.com"},{"url":"https://github.com/vllm-project/vllm/pull/44744","tags":["Issue Tracking","Patch"],"source":"security-advisories@github.com"},{"url":"https://github.com/vllm-project/vllm/security/advisories/GHSA-8wr5-jm2h-8r4f","tags":["Vendor Advisory","Exploit"],"source":"security-advisories@github.com"},{"url":"https://github.com/vllm-project/vllm/security/advisories/GHSA-8wr5-jm2h-8r4f","tags":["Vendor Advisory","Exploit"],"source":"134c704f-9b21-4f2e-91b3-4a467353bcc0"}],"sources":{"nvd":"https://nvd.nist.gov/vuln/detail/CVE-2026-54234","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."}