CVE-2026-63764: Server-Side Request Forgery (SSRF) in InternLM lmdeploy
A Server-Side Request Forgery (SSRF) vulnerability exists in InternLM's lmdeploy OpenAI-compatible API server. This flaw allows unauthenticated attackers to supply a crafted image_url that triggers HTTP 302 redirects to internal services or cloud metadata endpoints, bypassing URL safety checks. The vulnerability arises because redirects are followed without re-validating each redirect hop. This can lead to unauthorized internal resource access.
AI Analysis
Technical Summary
CVE-2026-63764 describes an SSRF vulnerability in the lmdeploy API server by InternLM. Attackers can send a POST request to the chat completions endpoint with an image_url parameter pointing initially to an attacker-controlled server. This server responds with an HTTP 302 redirect to internal network addresses such as loopback interfaces or cloud instance metadata endpoints. The server follows these redirects without reapplying URL safety validation, enabling attackers to access internal services and sensitive metadata endpoints without authentication.
Potential Impact
The vulnerability allows unauthenticated remote attackers to access internal network services and cloud metadata endpoints that are normally protected from external access. This can lead to information disclosure and potentially further compromise of the internal environment. The CVSS 4.0 score of 9.2 indicates a critical severity with network attack vector, no privileges or user interaction required, and high impact on confidentiality and integrity.
Mitigation Recommendations
Patch status is not yet confirmed — check the vendor advisory for current remediation guidance. Until an official fix is available, restrict access to the vulnerable API server and monitor for suspicious requests involving the image_url parameter. Consider implementing additional validation to prevent following unsafe redirects or disabling redirect following on the image_url fetch functionality.
CVE-2026-63764: Server-Side Request Forgery (SSRF) in InternLM lmdeploy
Description
A Server-Side Request Forgery (SSRF) vulnerability exists in InternLM's lmdeploy OpenAI-compatible API server. This flaw allows unauthenticated attackers to supply a crafted image_url that triggers HTTP 302 redirects to internal services or cloud metadata endpoints, bypassing URL safety checks. The vulnerability arises because redirects are followed without re-validating each redirect hop. This can lead to unauthorized internal resource access.
CVSS v4.0
Score 9.2critical
AI-Powered Analysis
Machine-generated threat intelligence
Technical Analysis
CVE-2026-63764 describes an SSRF vulnerability in the lmdeploy API server by InternLM. Attackers can send a POST request to the chat completions endpoint with an image_url parameter pointing initially to an attacker-controlled server. This server responds with an HTTP 302 redirect to internal network addresses such as loopback interfaces or cloud instance metadata endpoints. The server follows these redirects without reapplying URL safety validation, enabling attackers to access internal services and sensitive metadata endpoints without authentication.
Potential Impact
The vulnerability allows unauthenticated remote attackers to access internal network services and cloud metadata endpoints that are normally protected from external access. This can lead to information disclosure and potentially further compromise of the internal environment. The CVSS 4.0 score of 9.2 indicates a critical severity with network attack vector, no privileges or user interaction required, and high impact on confidentiality and integrity.
Mitigation Recommendations
Patch status is not yet confirmed — check the vendor advisory for current remediation guidance. Until an official fix is available, restrict access to the vulnerable API server and monitor for suspicious requests involving the image_url parameter. Consider implementing additional validation to prevent following unsafe redirects or disabling redirect following on the image_url fetch functionality.
Technical Details
- Data Version
- 5.2
- Assigner Short Name
- VulnCheck
- Date Reserved
- 2026-07-18T12:34:08.780Z
- Cvss Version
- 4.0
- State
- PUBLISHED
- Remediation Level
- null
Threat ID: 6a5fdc239c2644c7f8ba04d7
Added to database: 07/21/2026, 20:52:51 UTC
Last enriched: 07/21/2026, 21:07:57 UTC
Last updated: 07/21/2026, 22:01:12 UTC
Views: 5
Community Reviews
0 reviewsCrowdsource mitigation strategies, share intel context, and vote on the most helpful responses. Sign in to add your voice and help keep defenders ahead.
Want to contribute mitigation steps or threat intel context? Sign in or create an account to join the community discussion.
Actions
Updates to AI analysis require Pro Console access. Upgrade inside Console → Billing.
Need more coverage?
Upgrade to Pro Console for AI refresh and higher limits.
For incident response and remediation, OffSeq services can help resolve threats faster.
Latest Threats
Check if your credentials are on the dark web
Instant breach scanning across billions of leaked records. Free tier available.