CVE-2026-24779: vLLM MediaConnector vulnerable to Server-Side Request Forgery

# CVE-2026-24779: vLLM MediaConnector vulnerable to Server-Side Request Forgery

https://gcdradar.vercel.app/incidents/cve-2026-24779-vllm-mediaconnector-vulnerable-to-server-side · AIR-2026-0441 · closed 2026-01-27

- Lane: Patch now
- Severity: 2 (Disclosed vulnerability)
- Fix status: Fix status unknown
- First reported: 2026-01-27 (happened 2026-01-28)
- AI role: Attack on AI
- Products: vLLM
- CVE ids: CVE-2026-24779
- CVSS: 7.1

## Summary

A server-side request forgery vulnerability exists in vLLM due to inconsistent parsing of backslashes in URL handling libraries. Attackers can bypass host name restrictions to force the server to make requests to internal network resources.

## What to do

No action recorded; read the sources below and the vendor's advisory.

## Latest updates

- 2026-01-28: CVE-2026-24779: vLLM vulnerable to Server-Side Request Forgery (SSRF) through MediaConnector (https://github.com/advisories/GHSA-qh4c-xf7m-gxfc)
- 2026-01-27: CVE-2026-24779: vLLM is an inference and serving engine for large language models (LLMs). (https://nvd.nist.gov/vuln/detail/CVE-2026-24779)

## Sources

- 2026-01-27 NVD: CVE-2026-24779: vLLM is an inference and serving engine for large language models (LLMs). https://nvd.nist.gov/vuln/detail/CVE-2026-24779
- 2026-01-28 GitHub Advisory Database: CVE-2026-24779: vLLM vulnerable to Server-Side Request Forgery (SSRF) through MediaConnector https://github.com/advisories/GHSA-qh4c-xf7m-gxfc

Verify against the vendor's advisory before changing production. Data CC BY-SA 4.0, https://gcdradar.vercel.app.