CVE-2025-54381 BentoML is Vulnerable to an SSRF Attack Through File Upload Processing
Unauthenticated attackers can force BentoML servers to make arbitrary HTTP requests via file upload URLs, potentially accessing internal resources or cloud metadata.
- Exploited in the wild
- Not known
- Login needed
- Not needed
- Attackable
- Over the internet
- Victim action needed
- Not needed
- Fix
- Available
What is CVE-2025-54381?
CVE-2025-54381 is a critical server-side request forgery vulnerability in BentoML, published on 29 Jul 2025. An attacker can exploit it over the internet without logging in, with no action from the victim.
Official description
BentoML is a Python library for building online serving systems optimized for AI apps and model inference. In versions 1.4.0 until 1.4.19, the file upload processing system contains an SSRF vulnerability that allows unauthenticated remote attackers to force the server to make arbitrary HTTP requests. The vulnerability stems from the multipart form data and JSON request handlers, which automatically download files from user-provided URLs without validating whether those URLs point to internal network addresses, cloud metadata endpoints, or other restricted resources. The documentation explicitly promotes this URL-based file upload feature, making it an intended design that exposes all deployed services to SSRF attacks by default. Version 1.4.19 contains a patch for the issue.
Weakness: CWE-918
Which versions are affected by CVE-2025-54381?
3 open source packages ship an affected version, including bentoml. It affects 1 product.
| Package | Affected | Fixed in |
|---|---|---|
bentoml PyPI | From 1.4.0 | 1.4.19 |
bentoml PyPI | From 1.4.0 | 1.4.19 |
bentoml PyPI | From 1.4.0 | 1.4.19 |
Products include Bentoml Bentoml.
How do I fix CVE-2025-54381?
Upgrade bentoml to 1.4.19 or later.
Upgrade, and send outbound requests through an allow-list that blocks internal addresses. More about Server-side request forgery.
Is CVE-2025-54381 being exploited?
It is not confirmed as exploited in the wild, but public exploit code for CVE-2025-54381 exists.
Public exploit code: 1 proofs of concept on GitHub.
Other vulnerabilities in BentoML
- CVE-2025-32375 Insecure Deserialization leads to RCE in BentoML's runner server
- CVE-2025-27520 BentoML Allows Remote Code Execution (RCE) via Insecure Deserialization
- CVE-2026-44345 BentoML: Dockerfile command injection via docker.base_image
- CVE-2026-44346 BentoML: Dockerfile command injection via envs[*].name in bentofile.yaml
- CVE-2026-35044 BentoML has a Server-Side Template Injection via unsandboxed Jinja2 Environment in Dockerfile generation
Sources
Also known as GHSA-mrmq-3q62-6cc8, PYSEC-2026-297.