About Connect Custom APIs

[lhx20011226]

I have a Qianwen large language model (LLM) deployed on an internal Linux server. After adding it to “Connect Custom APIs,” it can only call the deepseek deployed locally on my Windows 10 computer via ollama, and it can't be called from anywhere else. What should I do?

[scosman]

I don't believe you have the server connected correctly. It should be "http://10.8.10.9:8020/v1" not "http://10.8.10.9:8020/v1/chat/co…"

After that it should work (assuming the URL and model name are correct)