CVE-2026-22778
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.
Scoring
- Severity
- CRITICAL
- CVSS base score
- 9.8
- CVSS vector
- CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H
- EPSS probability
- 3.72%
- CWE
- CWE-532
- Published
- 2026-02-02
- Last modified
- 2026-07-15
Affected products
- vllm-project vllm
Weakness type
Related vulnerabilities
- CVE-2026-79966 — CWE-532: Insertion of Sensitive Information into Log File
- CVE-2026-80169 — Dell SCG 5.0 Appliance versions prior to 5.36.00.16 and Dell SCG 5.0 Application versions prior to...
- CVE-2025-46808 — Sensitive information is leaked into NeuVector’s manager container logs
- CVE-2026-80124 — Dell SCG 5.0 Appliance versions prior to 5.36.00.16 and Dell SCG 5.0 Application versions prior to...
- CVE-2026-78631 — Improper Restriction of Sensitive Information in Okta Hyperdrive Agent Logging
- CVE-2026-78627 — Improper Credential Protection in Okta Hyperdrive Integration Installer Logging
- CVE-2026-68873 — Windows Program Compatibility Assistant Service Information Disclosure Vulnerability
- CVE-2026-86597 — Sensitive information written to logs by Snowflake drivers