CVE-2025-14921
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 · uneditedHugging Face Transformers Transformer-XL Model Deserialization of Untrusted Data Remote Code Execution Vulnerability. This vulnerability allows remote attackers to execute arbitrary code on affected installations of Hugging Face Transformers. User interaction is required to exploit this vulnerability in that the target must visit a malicious page or open a malicious file. The specific flaw exists within the parsing of model files. The issue results from the lack of proper validation of user-supplied data, which can result in deserialization of untrusted data. An attacker can leverage this vulnerability to execute code in the context of the current user. Was ZDI-CAN-25424.
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 confidenceThe vulnerability exists in Hugging Face Transformers' Transformer-XL model parsing logic. The lack of proper validation of user-supplied data during model file deserialization allows an attacker to craft malicious model files that, when loaded by a victim, execute arbitrary code via insecure deserialization. User interaction is required (opening a malicious model file).
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= 4.54.1CVSS 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
- Local
- Complexity
- Low
- Privileges
- None
- User interaction
- Required
- Scope
- Unchanged
- Confidentiality
- High
- Integrity
- High
- Availability
- High
CVSS:3.0/AV:L/AC:L/PR:N/UI:R/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 checksWork through these to decide whether this CVE applies to you.
-
Check installed Transformers versionRun 'pip show transformers' or 'import transformers; print(transformers.__version__)' to determine the installed versionAffected if Installed version is exactly 4.54.1
-
Verify Transformer-XL model loading codeSearch codebase for imports of 'transformers' and usage of 'TransfoXL' or 'TransformerXL' model classes, particularly any calls to from_pretrained() with Transformer-XL specific model namesAffected if Code loads or deserializes Transformer-XL model files using Transformers library
-
Identify model file sourcesReview code that loads model files - check for from_pretrained() calls, model loading functions, or configuration that specifies model paths from untrusted locationsAffected if Models are loaded from user-supplied paths, untrusted directories, or external sources without validation
-
Inspect model deserialization behaviorReview how model files are processed - check for any custom pickle loading, unsafe deserialization, or lack of file validation before loadingAffected if Model files are loaded without validation of file contents or signature verification prior to deserialization
You are affected if you run Transformers version 4.54.1 and load Transformer-XL model files, especially from untrusted sources, without validating file contents before deserialization.
Generated from the published advisory. Verify against your own configuration.
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.
From vendor dataUpgrade to a patched version of Hugging Face Transformers that implements proper validation of model file contents before deserialization, or implement model file validation/sandboxing as a temporary mitigation until the library is updated.
- Consultation4.0 h
- Implementation12.0 h
- Testing6.0 h
- Review / QA3.0 h
An estimate, not a bill — we confirm scope with you before any work starts. Need it this week? Rush from $7,040.
Scan for this in your stack
Free · runs locallyCheck whether your project pulls in CVE-2025-14921 — or any other known-vulnerable package — straight from your lock files. Free and open source; it runs locally and uploads nothing.
References Go to the primary sourcePrimary sources — vendor advisories, patches and trackers. Where our summary and a reference disagree, the reference wins.
Primary sourcesPractitioner notes
ContributedPeer-ranked notes from engineers who’ve handled CVE-2025-14921 in production — separate from our analysis above.
The advisory tells you what broke. It rarely tells you what actually worked. If you’ve dealt with this one, that detail is what the next engineer is searching for.
- The version that genuinely resolved it — not the one the vendor claimed
- A config change or rule that shut the vector down
- A gotcha in the upgrade path that cost you an afternoon
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.
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.
- 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