CVE-2026-22778
CWE-532Published: February 2, 2026· Updated: Jul 15, 2026
Official 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.
Risk Analysis
vLLM leaks heap addresses when processing invalid images, which reduces the effectiveness of ASLR. This can be chained with other vulnerabilities to achieve remote code execution.
No public exploit is known, and it is not in the KEV catalog. It is remotely exploitable with low complexity.
Upgrade vLLM to version 0.14.1 or later.
Technical Analysis
CVE-2026-22778 can be exploited remotely over the network without requiring physical or adjacent access, significantly expanding the attack surface for threat actors.
The vulnerability requires no privileges and no user interaction, making it a prime target for automated exploitation campaigns and worm-like propagation.
A successful exploit results in complete confidentiality breach (data exposure), full integrity compromise (data manipulation), availability disruption (denial of service), with a CVSS base score of 9.8.
CVSS v3.1 Vector Breakdown
Affected Vendors & Products
Exploit & PoC Resources
Official Patches & Advisories
All References (15)
Quick Facts
Related CVEs (CWE-532)
Recommended Actions
- →Apply vendor patches immediately
- →Monitor CVE-2026-22778 in threat intel feeds
- →Review IDS/IPS signatures for exploitation attempts