CVE-2026-105755
MEDIUM
4,2
Source: [email protected]
Attack Vector: network
Attack Complexity: high
Privileges Required: low
User Interaction: none
Scope: unchanged
Confidentiality: none
Integrity: low
Availability: low
Description
AI Translation Available
vLLM is an inference and serving engine for large language models. Prior to 0.30.0, flash late-interaction scoring at the /score and /rerank endpoints derives each worker's query_key value from the caller-controlled X-Request-Id header. A concurrent request that reuses a victim's identifier can overwrite the cached query embedding so the victim's documents are scored against the attacker's query, and shared use counters can also cause a late-interaction cache-miss error. This issue is fixed in version 0.30.0.
639
Authorization Bypass Through User-Controlled Key
IncompleteCommon Consequences
Security Scopes Affected:
Access Control
Potential Impacts:
Bypass Protection Mechanism
Gain Privileges Or Assume Identity
Applicable Platforms
All platforms may be affected
https://github.com/vllm-project/vllm/commit/ee17d0d869203ef9a35ad73358a4987bba1…
https://github.com/vllm-project/vllm/pull/51445
https://github.com/vllm-project/vllm/releases/tag/v0.30.0
https://github.com/vllm-project/vllm/security/advisories/GHSA-2phq-3phc-84px