Vulnerability in Vllm
CVE-2026-93840
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.
Published · last modified .
CVSS v3 metric
CVSS v3 base score 3.7 (Low). Vector: CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:L/A:N.
EPSS exploit prediction
EPSS: 0.002 (15.9th percentile), scored .
Very low probability of exploitation in the next 30 days; routine patching cadence is appropriate. 16th percentile — 15.9% of CVEs in the catalogue have a lower EPSS than this one. How to read EPSS.
EPSS over last 30 days · oldest: 0.002 · newest: 0.002 · change: 0.000
Affected products
Weakness classification (CWE)
References
- github.com/vllm-project/vllm/pull/49080 (issue-tracking, Patch, patch, Issue Tracking)
- github.com/vllm-project/vllm (Product, product)
- github.com/vllm-project/vllm/commit/5b0e5b69ac1a3884a6479c9537789c95263cc804 (Patch, patch)
- github.com/vllm-project/vllm/blob/v0.28.0/vllm/sampling_params.py (Product, technical-description)
- github.com/vllm-project/vllm/blob/v0.28.0/vllm/v1/worker/gpu/sample/logit_bias… (Product, technical-description)
- www.vulncheck.com/advisories/vllm-before-0.29.0-cross-request-logits-corruption… (Third Party Advisory, third-party-advisory)
Frequently asked questions
- What is CVE-2026-93840?
- CVE-2026-93840 is a low-severity vulnerability in Vllm, classified under Improper Validation of Array Index. CVSS score: 3.7/10. Published 2026-09-18.
- How severe is CVE-2026-93840?
- Low severity. CVSS v3 base score is 3.7 out of 10.