CVE-2025-46560: vLLM multimodal tokenizer quadratic time complexity resource exhaustion

Summary

vLLM versions from 0.8.0 up to but not including 0.8.5 contain a performance vulnerability in the input preprocessing logic of the multimodal tokenizer. Placeholder tokens such as <|audio_|> and <|image_|> are replaced through inefficient list concatenation, giving the algorithm quadratic time complexity (O(n²)). A malicious actor can send specially crafted inputs to trigger resource exhaustion.

Mitigation

Fixed in 0.8.5.

Affected packages in the Exposure Registry

Matched by package name and ecosystem. Each entry shows whether the package delegates to a language model and how many tracked packages depend on it.

• vllmPyPILLM dependency since 2023-08-25 · 15 tracked dependents

Topics

• Inference infrastructure

Related items

• LowCloakFEM: Agentic FEM Code Generation for Red-Teaming Computational Resource ExhaustionSimilar attack · Arxiv (cs.CR + cs.CL + cs.LG)
• MediumThe Rise of LLMJacking: How Threat Actors Monetize Compromised Cloud and Model ResourcesSimilar attack · Google Threat Intelligence
• InfoThe Blind Spots of Detection: Adversarial Perturbation Attacks against Watchdogs in Distributed LLM InferenceSimilar attack · Arxiv (cs.CR + cs.CL + cs.LG)
• HighCVE-2026-118761: Kubeflow Pipelines decompression bomb in pipeline archive uploadSimilar attack · NVD/CVE Database
• LowOpenCode User Hit with Hostile Backlash After Reporting Free AI LoopholeSimilar attack · Socket Blog