- CVSS
- MEDIUM · 6.5v3.1CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H
- Published
- 2026-08-13
- Weakness
- CWE-400
- Source
- nvd.nist.gov/vuln/detail/CVE-2026-73559
Description
vLLM is an inference and serving engine for large language models. From 0.19.0 until 0.26.0, the /v1/completions CompletionRequest.prompt field in vllm/entrypoints/openai/completion/protocol.py accepts an unbounded list[str] or list[list[int]], prompt_to_seq() in vllm/renderers/inputs/preprocess.py and OnlineRenderer.preprocess_completion() in vllm/renderers/online_renderer.py expand every element, and vllm/entrypoints/openai/completion/serving.py creates one engine generator and response slot per prompt, allowing an authenticated API client to exhaust CPU, memory, async scheduling capacity, engine request slots, and response buffering with one request. This issue is fixed in version 0.26.0.
References
- https://github.com/vllm-project/vllm/commit/675f4295cdfe0d870471c2b51bfeca3a68a9569e
- https://github.com/vllm-project/vllm/pull/47845
- https://github.com/vllm-project/vllm/releases/tag/v0.26.0
- https://github.com/vllm-project/vllm/security/advisories/GHSA-87x5-vmc3-756j
- https://github.com/vllm-project/vllm/security/advisories/GHSA-87x5-vmc3-756j
How GTK Cyber trains on this
AI security training at GTK Cyber covers the LLM and ML-pipeline vulnerability classes that vulnerabilities like CVE-2026-73559 fall into. Our hands-on courses are taught by Charles Givre and other practitioners who break and defend production AI systems.