Code execution via ignored remote-code opt-out
Published Sep 12, 2026 · Updated Sep 12, 2026
Code injection in vLLM before 0.28.0 allows attackers to execute arbitrary Python by loading a crafted LlavaOnevision2 model. The _load_ov2_processor function passes trust_remote_code to get_class_from_dynamic_module, whose catch-all keyword argument silently discards it before importing attacker-controlled processor modules. Execution requires an operator to load or serve the crafted model and runs top-level module code with the vLLM process or container's authority.
Summary
What happened
Code injection in vLLM before 0.28.0 allows attackers to execute arbitrary Python by loading a crafted LlavaOnevision2 model. The _load_ov2_processor function passes trust_remote_code to get_class_from_dynamic_module, whose catch-all keyword argument silently discards it before importing attacker-controlled processor modules. Execution requires an operator to load or serve the crafted model and runs top-level module code with the vLLM process or container's authority.
The record
- CVE
- CVE-2026-90553
- Published
- Sep 12, 2026
- Updated
- Sep 12, 2026
- Vendor
- vLLM Project
- Product
- vLLM
- Classifications
- CWE-829, CWE-94, T1059.006, T1204.002
- Attack vector
- local
- Privileges
- unauthenticated
Timeline
How it unfolded
- Sep 12, 2026CVE publishedPublication date reported by the CVE source.
- Sep 12, 2026Record updatedLatest update available in the CVE record.
Exploitability
Present is not the same as exploitable
Compare your product and version with the public record. A matching version still requires validation against your environment.
Is a vulnerable build present?
Compare these published version ranges with your installed build and any vendor patches.
- Affected versionversion=0 <0.28.0
What conditions does exploitation require?
What is affected?
Published CVSS scores
CVSS describes severity. EPSS estimates exploitation probability.
Attacks
What attackers are doing with it
Daily unique IPs observed by Shadowserver honeypots for known exploited vulnerabilities (KEVs). Missing observations do not establish an absence of attacks.
Weakness, pattern, technique
Public exploit references
No public exploit references are available in this record.
Labels summarize the accepted research assessment. They do not indicate a test against your environment.
Technologies
Your stack
See the directory against your own environment.
Your stack
Check the software in your environment
Book a demo to see how Hinoki identifies affected software and validates exploitability in your environment.
Book a demo