CVE-2025-46560
Versions starting from 0.8.0 and prior to 0.8.5 are affected by a critical performance vulnerability in the input preprocessing logic of the multimodal tokenizer.
Does this matter?
High impact if exploited, but EPSS currently rates exploitation as unlikely (0.50%). Schedule it in the normal patch cycle and watch for a rise in EPSS or a public exploit.
Description
vLLM is a high-throughput and memory-efficient inference and serving engine for LLMs. Versions starting from 0.8.0 and prior to 0.8.5 are affected by a critical performance vulnerability in the input preprocessing logic of the multimodal tokenizer. The code dynamically replaces placeholder tokens (e.g., <|audio_|>, <|image_|>) with repeated tokens based on precomputed lengths. Due to inefficient list concatenation operations, the algorithm exhibits quadratic time complexity (O(n²)), allowing malicious actors to trigger resource exhaustion via specially crafted inputs. This issue has been patched in version 0.8.5.
- CVSS 3.1
- 7.5 HIGHCVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H
- EPSS
- 0.50% probability · 41th percentile
- CISA KEV
- Not listed
- Weakness
- CWE-1333
- Affected
- vllm/vllm
- Source
- security-advisories@github.com
References
- https://github.com/vllm-project/vllm/blob/8cac35ba435906fb7eb07e44fe1a8c26e8744f4e/vllm/model_executor/models/phi4mm.py#L1182-L1197Product
- https://github.com/vllm-project/vllm/security/advisories/GHSA-vc6m-hm49-g9qgExploit, Vendor Advisory
- https://github.com/vllm-project/vllm/security/advisories/GHSA-vc6m-hm49-g9qgExploit, Vendor Advisory
Source: NVD record, EPSS from FIRST.org, KEV from CISA. Refreshed daily.