Vulnerability · Published Jul 7, 2026 · Assigned by VulnCheck

CVE-2026-59707LocalAI - Server-Side Request Forgery via POST /models/apply

High 8.6

In plain English

A high-severity server-side request forgery in LocalAI from LocalAI. Proof-of-concept exploit code is public. The chance it is exploited in the next 30 days is estimated at 1.4% (higher than 72% of all known vulnerabilities). Ready-made attack tooling exists: Nuclei. Fixed versions are available - upgrade the affected package.

Severity (CVSS 3.1)8.6High

Exploit odds (30 days)1.4%top 27.7% of all CVEs

Exploited in the wildNo reportsCISA SSVC: poc

Attack tooling1Nuclei

Official description

LocalAI contains an unauthenticated server-side request forgery vulnerability in the POST /models/apply endpoint that allows attackers to fetch arbitrary internal URLs. The endpoint passes unsanitized gallery URL fields directly to gallery.GetGalleryConfigFromURLWithContext without proper validation, enabling attackers to force the server to issue HTTP GET requests to private and loopback ranges with partial response content leaked through error messages.

Affected packages & fixed versions

Ecosystem	Package	Fixed in	Advisory
Go	github.com/mudler/LocalAI	1.40.1-0.20260703213242-2cbb3c96b34d	GHSA-8c5q-hx4g-qq23
Go	github.com/mudler/LocalAI	1.40.1-0.20260703213242-2cbb3c96b34d	GO-2026-6645

Affected products (as reported)

Vendor	Product	Versions
LocalAI	LocalAI	0 – through v4.3.1

Weakness: CWE-918 · Vector: CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:C/C:H/I:N/A:N

References

GitHub Issue issue-tracking
Product product
Patch Commit patch
https://www.vulncheck.com/advisories/localai-server-side-request-forgery-via-post-models-apply third-party-advisory