Low severity3.7NVD Advisory· Published Sep 18, 2026· Updated Sep 28, 2026
CVE-2026-93841
CVE-2026-93841
Description
vLLM through 0.29.0 contains a memory corruption vulnerability in the Triton _bincount_kernel where prompt token IDs index the penalty prompt-presence bitset without bounds checking against vocabulary size. Attackers can submit multimodal audio requests with tokens equal to vocabulary size, causing out-of-bounds writes that corrupt concurrent requests' sampler state and alter repetition penalty behavior.
AI Insight
LLM-synthesized narrative grounded in this CVE's description and references.
Affected products
3Patches
Vulnerability mechanics
References
4- github.com/vllm-project/vllm/pull/49081nvdIssue TrackingPatch
- www.vulncheck.com/advisories/vllm-through-0.29.0-adjacent-request-sampler-state-corruption-via-unvalidated-prompt-token-idsnvdThird Party Advisory
- github.com/vllm-project/vllm/blob/v0.29.0/vllm/v1/worker/gpu/sample/penalties.pynvdProduct
- github.com/vllm-project/vllm/blob/v0.29.0/vllm/v1/worker/gpu/sample/penalties.pynvdProduct
News mentions
1- VLLM: Batch of 17 Vulnerabilities Disclosed, Affecting LLM Inference Engine StabilityVypr Intelligence · Sep 21, 2026