- CVSS
- HIGH · 7.5v3.1CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H
- Published
- 2026-09-21
- Weakness
- CWE-617
- Source
- nvd.nist.gov/vuln/detail/CVE-2026-94623
Description
vLLM through 0.29.0 contains a denial of service vulnerability in the NIXL connector’s prefix caching implementation that fails to properly validate block counts across multi-prompt completion requests in prefill/decode disaggregated deployments. Attackers can trigger an assertion failure in NixlBaseConnectorWorker._apply_prefix_caching by submitting completion requests with multiple prompts of varying lengths, causing the decode worker to terminate and become unavailable until restarted.
References
- https://github.com/vllm-project/vllm
- https://github.com/vllm-project/vllm/blob/v0.29.0/vllm/distributed/kv_transfer/kv_connector/v1/nixl/base_worker.py#L2642-L2646
- https://github.com/vllm-project/vllm/pull/51505
- https://www.vulncheck.com/advisories/vllm-through-0.29.0-denial-of-service-via-nixl-multi-prompt-assertion-failure
How GTK Cyber trains on this
AI security training at GTK Cyber covers the LLM and ML-pipeline vulnerability classes that vulnerabilities like CVE-2026-94623 fall into. Our hands-on courses are taught by Charles Givre and other practitioners who break and defend production AI systems.