CVE-2026-93840
vLLM before 0.29.0 Cross-Request Logits Corruption via allowed_token_ids
CVSS Score
3.7
EPSS Score
0.0%
EPSS Percentile
0th
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.
| CWE | CWE-129 |
| Vendor | vllm-project |
| Product | vllm |
| Published | Sep 18, 2026 |
Stay Ahead of the Next One
Get instant alerts for vllm-project vllm
Be the first to know when new low vulnerabilities affecting vllm-project vllm are published — delivered to Slack, Telegram or Discord.
Get Free Alerts →
Free · No credit card · 60 sec setup
CVSS v3 Breakdown
CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:L/A:N Attack Vector
Network
Attack Complexity
High
Privileges Required
None
User Interaction
None
Scope
Unchanged
Confidentiality
None
Integrity
Low
Availability
None
Affected Versions
vllm-project / vllm
0 < 0.29.0
References
github.com: https://github.com/vllm-project/vllm/pull/49080 github.com: https://github.com/vllm-project/vllm github.com: https://github.com/vllm-project/vllm/commit/5b0e5b69ac1a3884a6479c9537789c95263cc804 github.com: https://github.com/vllm-project/vllm/blob/v0.28.0/vllm/sampling_params.py#L881-L903 github.com: https://github.com/vllm-project/vllm/blob/v0.28.0/vllm/v1/worker/gpu/sample/logit_bias.py#L179-L191 vulncheck.com: https://www.vulncheck.com/advisories/vllm-before-0.29.0-cross-request-logits-corruption-via-allowed-token-ids
Credits
🔍 Mingkai Yu 🔍 Jiapeng Li 🔍 Jiajia Liu