CVE-2026-34755: Vllm

Medium severity, CVSS 6.5. EPSS: 0.8% chance of exploitation in the next 30 days.

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.

Affected products

• Vllm Vllm: from 0.7.0, before 0.19.0 (fixed in 0.19.0)

Published 2026-04-06. Last modified 2026-08-25.