dbcve.org
VllmApplication

CVE-2026-22807

CRITICAL · 9.8 CVSS v3.1 Published 2026-01-21
Fix available
A fix is available. Upgrade to 0.14.0 or later.
See remediation →
100/100
Remediation priority · Urgent
Remotely reachable No privileges Zero-click Patch available

Official description Straight from the sourceThe vendor's or NVD's own wording, published unedited. Authoritative, but often terse — it says what broke, rarely what to do.

NVD · unedited
vLLM is an inference and serving engine for large language models (LLMs). Starting in version 0.10.1 and prior to version 0.14.0, 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. 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. Version 0.14.0 fixes the issue.

Technical summary Written by usOur analysis, written from the advisory, the CVSS vector and the affected-version data. It adds context the advisory leaves out, and never invents facts that are not in the source.

dbcve analysis · high confidence

vLLM versions 0.10.1 through prior to 0.14.0 load Hugging Face `auto_map` dynamic modules during model resolution without checking the `trust_remote_code` setting. This allows an attacker who can control or influence the model repository (local directory or remote Hugging Face repo) to inject arbitrary Python code that executes on the vLLM server at startup, before any API requests are handled.

MitigationUpgrade vLLM to version 0.14.0 or later. Additionally, audit all model repositories and paths currently loaded by vLLM deployments to ensure no untrusted or compromised models were introduced, and implement controls to validate model sources before loading.

Verify against the referenced sources before acting — the references below are authoritative for this CVE, this summary is not.

Affected products & versions What the vendor confirmedThe version ranges the vendor confirmed as vulnerable. If your version sits inside a range here, treat yourself as exposed until you have upgraded.

NVD · CPE data
VllmApplication
Affected:>= 0.10.1, < 0.14.0

CVSS breakdown How the score is builtThe industry scoring standard. It rates how the flaw is reached, what it takes to exploit, and what an attacker gains — the score is derived from those, not the other way round.

From the vector
Attack vector
Network
Complexity
Low
Privileges
None
User interaction
None
Scope
Unchanged
Confidentiality
High
Integrity
High
Availability
High

CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H

Am I affected? How to checkSteps we derive from the advisory and the affected-version data, so you can decide whether this CVE reaches your setup. They are a guide, not a scan — your own configuration is the authority.

dbcve checks

Work through these to decide whether this CVE applies to you.

  1. Check installed vLLM version
    Run `pip show vllm` or `python -c "import vllm; print(vllm.__version__)"` to retrieve the installed version number
    Affected if Version is 0.10.1 or higher but lower than 0.14.0
  2. Identify model configurations using auto_map
    Search model config files (config.json) and deployment configs for the 'auto_map' key, which triggers dynamic module loading during model resolution
    Affected if Any model configuration contains the 'auto_map' parameter pointing to custom model classes
  3. Review trust_remote_code setting in model configs
    Inspect config.json files in model directories or vLLM serving configurations for the trust_remote_code flag
    Affected if trust_remote_code is not explicitly set to False, or auto_map is used without verifying this setting is disabled
  4. Audit configured model repositories and paths
    Examine --model, --hf_model_config, or similar startup parameters and config files that specify where vLLM loads models from
    Affected if Models are loaded from untrusted or attacker-controlled repositories or paths, or any path that permits world-writable access

You are affected if vLLM version is between 0.10.1 and 0.13.x inclusive and any configured model uses auto_map to load dynamic modules, especially from untrusted model sources.

Generated from the published advisory. Verify against your own configuration.

Check your environment

Paste your version and any relevant configuration and it will be compared against the affected criteria above. Do not include secrets or credentials.

AI-assisted, checked against the advisory. Informational, not a guarantee.

Remediation Closing itWhat it takes to close this. Where a vendor fix exists we point at it; where none exists we say so plainly, and can build one. Effort estimates are scoped from the advisory, not from your codebase.

dbcve · scoped
Upgrade available Upgrade to 0.14.0 or later
Fixed in 0.14.0
Vendor patch github.com →
Interim mitigation

Upgrade vLLM to version 0.14.0 or later. Additionally, audit all model repositories and paths currently loaded by vLLM deployments to ensure no untrusted or compromised models were introduced, and implement controls to validate model sources before loading.

Recommended fix High confidence

0.14.0

  1. Upgrade vLLM to version 0.14.0 or later using your package manager (e.g., pip install vllm>=0.14.0)
  2. After upgrading, verify the installation: python -c "import vllm; print(vllm.__version__)"
  3. Ensure any automated model loading processes use the updated vLLM version
  4. If using containerized deployments, rebuild containers with the updated vLLM version

Generated from the published advisory — verify against the referenced sources before acting.

Fix this in Vllm Scoped from the published advisory
  • Consultation4.0 h
  • Implementation2.0 h
  • Testing4.0 h
  • Review / QA2.0 h
12.0 hours of engineering $2,120
Get the upgrade done

An estimate, not a bill — we confirm scope with you before any work starts. Need it this week? Rush from $3,392.

Scan for this in your stack

Free · runs locally
dbcve dependency scanner

Check whether your project pulls in CVE-2026-22807 — or any other known-vulnerable package — straight from your lock files. Free and open source; it runs locally and uploads nothing.

Get it on GitHub Scan in the browser

References Go to the primary sourcePrimary sources — vendor advisories, patches and trackers. Where our summary and a reference disagree, the reference wins.

Primary sources

Practitioner notes

Contributed

Peer-ranked notes from engineers who’ve handled CVE-2026-22807 in production — separate from our analysis above.

No notes yet

Be the first to add a field note for this CVE — a mitigation you’ve verified, a version caveat, or a link to a working fix. Sign in above to contribute.

What this is

A place for practitioners to share what actually worked: a mitigation you’ve tested, a configuration change, a version- or environment-specific caveat, or a link to a verified patch. The most useful notes rise to the top as peers upvote them, so the signal stays high.

What belongs here
  • Verified mitigations, workarounds, and config changes
  • Version or environment caveats, and links to real fixes
  • No weaponised exploit code, or anything meant to cause harm
  • No spam, self-promotion, credentials, or personal data