CVE-2026-22778

vLLM leaks a heap address when PIL throws an error

State
Published
Severity
9.8 Critical
Published
Feb 2, 2026
Updated
Jul 15, 2026
Reserved
Jan 9, 2026

Read the record on cve.org

Description

vLLM is an inference and serving engine for large language models (LLMs). From 0.8.3 to before 0.14.1, when an invalid image is sent to vLLM's multimodal endpoint, PIL throws an error. vLLM returns this error to the client, leaking a heap address. With this leak, we reduce ASLR from 4 billion guesses to ~8 guesses. This vulnerability can be chained a heap overflow with JPEG2000 decoder in OpenCV/FFmpeg to achieve remote code execution. This vulnerability is fixed in 0.14.1.

Scores

CVSSScoreVectorSource
3.19.8 CriticalCVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:HGitHub_M
3.19.8 CriticalCVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:Hredhat-SADP

Weaknesses

  • CWE-532 Insertion of Sensitive Information into Log File
  • CWE-209 Generation of Error Message Containing Sensitive Information

Affected Products

ProductVersions
vllmvllm-project
  • >= 0.8.3, < 0.14.1Affected

References

Every record here comes from the CVE Program’s CVE List. © 1999–2026 The MITRE Corporation. CVE is a registered trademark of The MITRE Corporation. The records are reproduced under the CVE Program terms of use.