vLLM: VLLM before 0.29.0 validates allowed_token_ids against tokenizer length instead of model output…
vLLM
More for vLLM RSS feed for vLLM
- Severity
- Medium, CVSS 5.3
- CVSS vector
- CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:L/A:N
- Record
- CVE-2026-93840
- Affects
- vLLM
- Known exploited
- Not listed in the CISA Known Exploited Vulnerabilities catalog
- Published
- Record first published by the source
- September 18, 2026
- Source last modified
- September 28, 2026
Description
vLLM before 0.29.0 validates allowed_token_ids against tokenizer length instead of model output logits width in SamplingParams._validate_allowed_token_ids(). Attackers can supply token IDs above the output vocabulary that pass validation, causing LogitBiasState to corrupt GPU logits state and allow concurrent requests to sample tokens outside their allowlists.
Description quoted from the NVD record.