Description
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.
Affected products
vllm: vllm
References4
- github.com/vllm-project/vllm/security/advisories/GHSA-2phq-3phc… Vendor Advisory
- github.com/vllm-project/vllm/commit/ee17d0d869203ef9a35ad73358a… Patch
- github.com/vllm-project/vllm/pull/51445 Issue TrackingPatch
- github.com/vllm-project/vllm/releases/tag/v0.30.0 Release Notes
Metrics
- CVSS 3.1
-
4.2
MEDIUM
Score source: CNA
CVSS:3.1/AV:N/AC:H/PR:L/UI:N/S:U/C:N/I:L/A:L - EPSS — exploit probability
- 0% percentile: 10 · as of 10 November 2026
- CWE
- CWE-639
Could this be exploited in your environment?
A patch closes the known hole. A penetration test shows what an attacker can still reach — with proof.