LightLLM versions up to 1.2.0 do not properly validate the image_url and audio_url parameters in its multimodal endpoints. This flaw allows unauthenticated attackers to perform SSRF attacks by supplying arbitrary URLs. The server processes these URLs, which may result in fetching internal resources. The vision model processing can disclose the content of these resources or error messages that reveal internal network structure. The CVSS 4.0 base score is 6.9, indicating a medium severity vulnerability with network attack vector, no privileges or user interaction required, and low impact on confidentiality.

An attacker can exploit this vulnerability to make the server perform requests to internal or otherwise inaccessible resources. This can lead to information disclosure about internal network topology or content of internal resources. The vulnerability does not require authentication or user interaction, increasing its risk. However, the impact on confidentiality is rated low, and there is no impact on integrity or availability.

No vendor advisory or patch information is provided. Patch status is not yet confirmed — check the vendor advisory for current remediation guidance. Until a fix is available, restrict access to the vulnerable endpoints and monitor for unusual request patterns targeting image_url and audio_url parameters.