Tools AI Risk Radar ai-incident-0175
Incident record
vLLM prompt embeddings can crash shared inference
- Severity
- Medium
- Status
- Proof-of-concept
- Type
- Infra Vuln
- Target
- vLLM through 0.31.0 with prompt_embeds and penalties
- Actor
- researcher
- CVE
- CVE-2026-105922
What happened
A request combining supplied prompt embeddings with repetition or frequency penalties could stop the shared inference engine. The report covers vLLM through 0.31.0.
The original report was published on 19 September. The CVE rates it 4.3 under CVSS 3.1 and 5.3 under CVSS 4.0, both Medium. A linked proposed fix alone does not establish a released patch.
Sources
- Primary disclosuregithub.com/vllm-project/vllm/issues/57719
- Supporting advisoryraw.githubusercontent.com/CVEProject/cvelistV5/main/cves/202…
One record from the AI Risk Radar, maintained by the Security Desk.Data: CSV · JSON ·RSS · CC BY 4.0 with attribution to YFarmX.Tracker updated · 13 October 2026