CVSS 4.0
6.3 mediumCVSS:4.0/AV:N/AC:L/AT:P/PR:N/UI:N/VC:N/VI:L/VA:N/SC:N/SI:N/SA:NCVE-2026-93840
Medium (6.3) in vllm vllm. EPSS 0.2% chance of exploitation in the next 30 days. Fixed in vllm 0.29.0.
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.
Severity
CVSS 3.1
5.3 mediumCVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:L/A:N- EPSS
- 0.2% chance of exploitation in the next 30 days; higher than 15% of CVEs (scores of 2026-10-31)
- CWE
- CWE-129
- Published
- Modified
- Record from
- the NVD, synced
Affected
- vllm:vllm
Packages (OSV)
| Advisory | Ecosystem | Package | Affected | Fixed in |
|---|---|---|---|---|
| PYSEC-2026-3998 | PyPI | vllm | <0.29.0 | 0.29.0 |
References
- https://github.com/vllm-project/vllm
- https://github.com/vllm-project/vllm/blob/v0.28.0/vllm/sampling_params.py#L881-L903
- https://github.com/vllm-project/vllm/blob/v0.28.0/vllm/v1/worker/gpu/sample/logit_bias.py#L179-L191
- https://github.com/vllm-project/vllm/commit/5b0e5b69ac1a3884a6479c9537789c95263cc804
- https://github.com/vllm-project/vllm/pull/49080