VDB
CVE-2026-22778
CVE-2026-22778
PUBLISHED
CVSS 9.800000190734863 CRITICAL
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.
EPSS 0.09% · 25.0th percentile
Risk Scores
CVSS v3.1
9.800000190734863
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H
EPSS Score
0.09%
25.0th percentile
Affected Products
| Vendor | Product | Versions |
|---|---|---|
| vllm-project | vllm | >= 0.8.3, < 0.14.1, >= 0.8.3, < 0.14.1 |
| vllm | vllm | 0.8.3, 0.8.3 |
| PyPI | vllm | 0.8.3, 0.8.3 |
Timeline
- Jan 8, 2026 Fix PR Merged
- Jan 14, 2026 Fix PR Merged
- Feb 2, 2026 CVE Published
- Feb 2, 2026 PoC Published
- Feb 2, 2026 PoC Published
- Feb 2, 2026 PoC Published
- Feb 3, 2026 CVE Updated
- Feb 3, 2026 EPSS Score
- Feb 3, 2026 Security Advisory
- Feb 5, 2026 EPSS Score
- Feb 5, 2026 PoC Published
- Feb 7, 2026 EPSS Score
References
- https://github.com/vllm-project/vllm/security/advisories/GHSA-4r2x-xpjr-7cvv url
- https://github.com/vllm-project/vllm/pull/31987 url
- https://github.com/vllm-project/vllm/pull/32319 url
- https://github.com/vllm-project/vllm/releases/tag/v0.14.1 url
- https://nvd.nist.gov/vuln/detail/CVE-2026-22778 advisory
- https://github.com/vllm-project/vllm package