CVE-2026-33626 in the critical feed json
LMDeploy Vulnerable to Server-Side Request Forgery (SSRF) via Vision-Language Image Loading
published 2026-04-20 6 mo ago modified 2026-10-26 13 d ago also EUVD-2026-23970 , GHSA-6w67-hwm5-92mq , PYSEC-2026-2607
- cvss
- 7.5
- version 3.1, CNA
- cisa kev
- Not listed
- not in CISA's catalog
- epss
- 45.3%
- EPSS 45.3% chance of exploitation in 30 days, percentile 98%
- public exploits
- 0
- none known, plus a Nuclei check
Summary
LMDeploy is a toolkit for compressing, deploying, and serving large language models. Versions prior to 0.12.3 have a Server-Side Request Forgery (SSRF) vulnerability in LMDeploy's vision-language module. The `load_image()` function in `lmdeploy/vl/utils.py` fetches arbitrary URLs without validating internal/private IP addresses, allowing attackers to access cloud metadata services, internal networks, and sensitive resources. Version 0.12.3 patches the issue.
Patch / fix
- internlm lmdeploy 0.12.3
- github.com/InternLM/lmdeploy@71d64a3 commit
- github.com/InternLM/lmdeploy#4447 pull request
- github.com/InternLM/lmdeploy/releases/tag/v0.12.3 release
CISA SSVC
- Exploitation
- Proof of concept is public
- Automatable
- yes
- Technical impact
- partial
- Assessed
- 6 mo ago
Exploit prediction
45.25% chance of exploitation in the next 30 days, higher than 98% of all scored CVEs (FIRST EPSS, 2026-11-08).
Exploits and checks
- Nuclei template LMDeploy - Server-Side Request Forgery nuclei detection template
- github.com https://github.com/InternLM/lmdeploy/security/advisories/GHSA-6w67-hwm5-92mq referenced PoC, tagged "exploit" by a source, unverified
CVSS scores
| assigned by | version | score | vector |
|---|---|---|---|
| CNA | 3.1 | 7.5 | CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:N/A:N |
| GitHub | 3.1 | 7.5 | CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:N/A:N |
| OSV | 3.1 | 7.5 | CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:N/A:N |
Affected packages
| ecosystem | package | affected | fixed in |
|---|---|---|---|
| pypi | lmdeploy | <=0.12.2 |
no fix listed |
Affected products
- InternLM lmdeploy
Classification
- Weakness
-
- CWE-918 Server-Side Request Forgery
References
-
github.com
/InternLM/lmdeploy/pull/4447
web, issue-tracking, patch
-
github.com
/InternLM/lmdeploy/commit/71d64a339edb901e9005358e0633fbbab367d626
web, patch
-
github.com
/InternLM/lmdeploy/security/advisories/GHSA-6w67-hwm5-92mq
exploit, web, mitigation, vendor-advisory
-
github.com
/InternLM/lmdeploy/releases/tag/v0.12.3
web, release-notes
-
nvd.nist.gov
/vuln/detail/CVE-2026-33626
advisory
-
github.com
/InternLM/lmdeploy
package
-
pypi.org
/project/lmdeploy
package
-
github.com
/advisories/GHSA-6w67-hwm5-92mq
advisory
Sources
- CVE List CVE-2026-33626
- GitHub Advisory GHSA-6w67-hwm5-92mq
- NVD CVE-2026-33626
- OSV PYSEC-2026-2607
Fencer Vulns keeps one entry per vulnerability: the CVE id wins, then GHSA, and records that alias each other are merged, except that two CVEs always keep separate entries. The same data is available as JSON.
CPEs and fixed versions use Anchore's correction of NVD's data: Added CPE configurations because not yet analyzed by NVD.