VULNERABILITY INTELLIGENCE

CVE-2026-34755

CVSS score6.5 MEDIUM
EPSS probability0.46%
CISA KEVNot currently listed
Published2026-04-06
Last modified2026-08-25
CVSS vectorCVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H

Description

vLLM is an inference and serving engine for large language models (LLMs). From 0.7.0 to before 0.19.0, the VideoMediaIO.load_base64() method at vllm/multimodal/media/video.py splits video/jpeg data URLs by comma to extract individual JPEG frames, but does not enforce a frame count limit. The num_frames parameter (default: 32), which is enforced by the load_bytes() code path, is completely bypassed in the video/jpeg base64 path. An attacker can send a single API request containing thousands of comma-separated base64-encoded JPEG frames, causing the server to decode all frames into memory and crash with OOM. This vulnerability is fixed in 0.19.0.

Weakness classification

  • security-advisories@github.com: CWE-770
  • 0b0ca135-0b70-47e7-9f44-1890c2a1c46c: CWE-770

Primary references