HighVulnerabilityLLM-specific

CVE-2026-33626: LMDeploy SSRF in load_image() of vision-language module via arbitrary URLs

Source: NVD(opens in a new tab)
Identifier: CVE-2026-33626
Published: April 20, 2026
Record updated: October 9, 2026

Affected: lmdeploy <= 0.12.2
Fixed in: No fixed version was stated when the source was last read.
Known exploitation: Not listed in the CISA Known Exploited Vulnerabilities catalog at the last check.
EPSS: 1.5%

Summary

CVE-2026-33626 is a Server-Side Request Forgery (SSRF) flaw in LMDeploy versions prior to 0.12.3, within the vision-language module. The load_image() function in lmdeploy/vl/utils.py fetches arbitrary URLs without checking for internal or private IP addresses, letting an attacker reach cloud metadata services, internal networks and sensitive resources. The weakness is classified as CWE-918.

Mitigation

Version 0.12.3 patches the issue.

Details

Scores and exploitation

CVSS score: 7.5(high)
CVSS vector: CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:N/A:N
Attack vector: network
Attack complexity: low
Privileges required: none
User interaction: none
EPSS: 1.5%Estimated probability that this CVE is exploited in the next 30 days, from the FIRST Exploit Prediction Scoring System. Read on October 12, 2026.
Disclosure date: April 20, 2026

Weaknesses and techniques

Weakness (CWE): CWE-918
Attack pattern (CAPEC): CAPEC-664

Affected

Packages: lmdeploy@<= 0.12.2

Classification

Attack type: Other
Attack sophistication: Trivial
AI component targeted: Inference
Impact: Confidentiality
Labels: security

How this record was produced

Summaries, mitigations and labels on this page are generated automatically from the source. The linked source is authoritative. The source gives this advisory a description and no title, so the title above is a headline generated from the source text.

Original source: https://nvd.nist.gov/vuln/detail/CVE-2026-33626

First tracked: April 20, 2026 at 08:08 PM ET, from NVD/CVE Database

Classifier: claude-haiku-5-5, prompt v4

Classifier confidence: 95% (self-assessed)

Summary

Mitigation

Related items