CVE-2026-25960

Published Mar 9, 2026

Last updated a month ago

CVSS high 7.1
Business logic

Overview

Description
vLLM is an inference and serving engine for large language models (LLMs). The SSRF protection fix for CVE-2026-24779 add in 0.15.1 can be bypassed in the load_from_url_async method due to inconsistent URL parsing behavior between the validation layer and the actual HTTP client. The SSRF fix uses urllib3.util.parse_url() to validate and extract the hostname from user-provided URLs. However, load_from_url_async uses aiohttp for making the actual HTTP requests, and aiohttp internally uses the yarl library for URL parsing. This vulnerability in 0.17.0.
Source
security-advisories@github.com
NVD status
Analyzed
Products
vllm

Risk scores

CVSS 3.1

Type
Primary
Base score
9.8
Impact score
5.9
Exploitability score
3.9
Vector string
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H
Severity
CRITICAL

Weaknesses

security-advisories@github.com
CWE-918

Social media

Hype score
Not currently trending

Configurations

Related CVEs

  1. vLLM is an inference and serving engine for large language models (LLMs). From 0.16.0 to before 0.19.0, a server-side request forgery (SSRF) vulnerability in download_bytes_from_url allows any actor who can control batch input JSON to make the vLLM batch runner issue arbitrary HTTP/HTTPS requests from the server, without any URL validation or domain restrictions. This can be used to target internal services (e.g. cloud metadata endpoints or internal HTTP APIs) reachable from the vLLM host. This vulnerability is fixed in 0.19.0.•CVE-2026-34753
  2. vLLM is an inference and serving engine for large language models (LLMs). From 0.7.0 to before 0.19.0, the VideoMediaIO.load_base64() method at vllm/multimodal/media/video.py splits video/jpeg data URLs by comma to extract individual JPEG frames, but does not enforce a frame count limit. The num_frames parameter (default: 32), which is enforced by the load_bytes() code path, is completely bypassed in the video/jpeg base64 path. An attacker can send a single API request containing thousands of comma-separated base64-encoded JPEG frames, causing the server to decode all frames into memory and crash with OOM. This vulnerability is fixed in 0.19.0.•CVE-2026-34755
  3. vLLM is an inference and serving engine for large language models (LLMs). From 0.1.0 to before 0.19.0, a Denial of Service vulnerability exists in the vLLM OpenAI-compatible API server. Due to the lack of an upper bound validation on the n parameter in the ChatCompletionRequest and CompletionRequest Pydantic models, an unauthenticated attacker can send a single HTTP request with an astronomically large n value. This completely blocks the Python asyncio event loop and causes immediate Out-Of-Memory crashes by allocating millions of request object copies in the heap before the request even reaches the scheduling queue. This vulnerability is fixed in 0.19.0.•CVE-2026-34756
  4. A Business Logic vulnerability exists in SourceCodester Pharmacy Product Management System 1.0. The vulnerability is located in the add-sales.php file. The application fails to validate the "txtprice" and "txttotalcost" parameters, allowing attackers to submit negative values for sales transactions. This leads to incorrect financial calculations, corruption of sales reports, and potential financial loss.•CVE-2026-30573
  5. A Business Logic vulnerability exists in SourceCodester Loan Management System v1.0 due to the lack of proper input validation. The application allows administrators to define "Loan Plans" which determine the duration of a loan (in months). However, the backend fails to validate that the duration must be a positive integer. An attacker can submit a negative value for the months parameter. The system accepts this invalid data and creates a loan plan with a negative duration.•CVE-2026-30523

References

Sources include official advisories and independent security research.