Remote code execution via hardcoded model trust
Published May 28, 2026 · Updated May 28, 2026
Remote code execution in vLLM 0.14.1 allows attackers to run Python code when operators load malicious Hugging Face model repositories. The NemotronVL vision-model initializer and KimiK25 image-processor loader force trust_remote_code=True instead of honoring the configured opt-out. Exploitation requires an operator to load an attacker-controlled repository, after which its Python code executes with the vLLM process permissions.
Summary
What happened
Remote code execution in vLLM 0.14.1 allows attackers to run Python code when operators load malicious Hugging Face model repositories. The NemotronVL vision-model initializer and KimiK25 image-processor loader force trust_remote_code=True instead of honoring the configured opt-out. Exploitation requires an operator to load an attacker-controlled repository, after which its Python code executes with the vLLM process permissions.
The record
- CVE
- CVE-2026-4944
- Published
- May 28, 2026
- Updated
- May 28, 2026
- Vendor
- vLLM Project
- Product
- vLLM
- Classifications
- CWE-22, T1203
- Attack vector
- network
- Privileges
- unauthenticated
Timeline
How it unfolded
- May 28, 2026CVE publishedPublication date reported by the CVE source.
- May 28, 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=unspecified <=latest
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
- Huntr malicious-model repository proof of conceptproof of concept · unverified
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