vLLM affected by RCE via auto_map dynamic module loading during model initialization
Published 21 Jan 2026 · Updated 10 Sept 2026
Is your project on an affected vllm?
Look after several Python apps? Indie watches up to 15 repos, private ones too, for $5 a month. See Indie
Affected package
- Affected
- >=0.10.1 <0.14.0
- Fixed in
- 0.14.0
pip install 'vllm>=0.14.0'Details
Summary
vLLM loads Hugging Face auto_map dynamic modules during model resolution without gating on trust_remote_code, allowing attacker-controlled Python code in a model repo/path to execute at server startup.
---
Impact
An attacker who can influence the model repo/path (local directory or remote Hugging Face repo) can achieve arbitrary code execution on the vLLM host during model load. This happens before any request handling and does not require API access.
---
Affected Versions
All versions where vllm/model_executor/models/registry.py resolves auto_map entries with try_get_class_from_dynamic_module without checking trust_remote_code (at least current main).
---
Details
During model resolution, vLLM unconditionally iterates auto_map entries from the model config and calls try_get_class_from_dynamic_module, which delegates to Transformers’ get_class_from_dynamic_module and executes the module code.
This occurs even when trust_remote_code is false, allowing a malicious model repo to embed code in a referenced module and have it executed during initialization.
Relevant code
vllm/model_executor/models/registry.py:856— auto_map resolutionvllm/transformers_utils/dynamic_module.py:13— delegates toget_class_from_dynamic_module, which executes code
---
Fixes
- https://github.com/vllm-project/vllm/pull/32194
Credits
Reported by bugbunny.ai
Is GHSA-2pc9-4j83-qjmr in any of your client repos?
Checking one package by hand is fine. Checking it across forty client projects is not. NPMscan reads the lockfiles in every GitHub repo you pick, finds each project with an affected vllm, and gives you the command that fixes it.
More advisories for vllm
- LowCVE-2026-115917vLLM: Permissive default CORS origins allow cross-origin requests to the vLLM HTTP server · 28 Oct 2026
- MediumCVE-2026-115913vLLM: Phi-4 multimodal audio encoder materializes and retains full attention matrices: a single max-length audio OOM-crashes the vLLM engine · 28 Oct 2026
- MediumCVE-2026-115914vLLM DeepSeek-OCR image processing vulnerable to denial of service · 28 Oct 2026
- HighCVE-2026-115915vLLM: Untrusted EPS inputs can trigger denial of service or Ghostscript RCE · 28 Oct 2026
- HighCVE-2026-115912vLLM: Public multimodal SHM handle forgery reaches worker-side `pickle.loads()` with insecure serialization disabled · 28 Oct 2026
- MediumCVE-2026-115916vLLM: Out-of-vocabulary prompt token id passes input validation and triggers a CUDA device-side assert (engine crash / DoS) · 28 Oct 2026
References
- Linkgithub.com/vllm-project/vllm/security/advisories/GHSA-2pc9-4j83-qjmr
- Advisorynvd.nist.gov/vuln/detail/CVE-2026-22807
- Linkgithub.com/vllm-project/vllm/pull/32194
- Linkgithub.com/vllm-project/vllm/commit/78d13ea9de4b1ce5e4d8a5af9738fea71fb024e5
- Linksecurity.access.redhat.com/data/csaf/v2/vex/2026/cve-2026-22807.json
- Linkpypi.org/project/vllm
- Linkgithub.com/vllm-project/vllm/releases/tag/v0.14.0
- Packagegithub.com/vllm-project/vllm
- Linkgithub.com/pypa/advisory-database/tree/main/vulns/vllm/PYSEC-2026-2010.yaml
- Advisorygithub.com/advisories/GHSA-2pc9-4j83-qjmr
- Linkbugzilla.redhat.com/show_bug.cgi?id=2431865
- Linkaccess.redhat.com/security/cve/CVE-2026-22807
- Linkaccess.redhat.com/errata/RHSA-2026:5119
- Linkaccess.redhat.com/errata/RHSA-2026:42644
- Linkaccess.redhat.com/errata/RHSA-2026:3782
- Linkaccess.redhat.com/errata/RHSA-2026:3713
- Linkaccess.redhat.com/errata/RHSA-2026:3462
- Linkaccess.redhat.com/errata/RHSA-2026:3461
- Linkaccess.redhat.com/errata/RHSA-2026:30089
- Linkaccess.redhat.com/errata/RHSA-2026:30088
Advisory data from OSV.dev (opens OSV.dev in a new tab), refreshed daily.