Custom servers

Any server that speaks the OpenAI chat-completions API works with kunji: LM Studio, vLLM, LiteLLM, llama.cpp’s server, or a company proxy.

Add a server

kunji providers add lmstudio --base-url http://localhost:1234/v1 --no-key --name "LM Studio"
kunji keys add lmstudio
kunji chat -m lmstudio:<model>

• id (lmstudio): your short name, lowercase letters, numbers and dashes.
• --base-url: the address that ends in /v1 (the part before /chat/completions).
• --no-key: use this if the server doesn’t need a key. Otherwise kunji keys add <id> asks for one and stores it in the keychain.

From the menu: Providers and keys → Add a custom server.

Examples

Server	Base URL
LM Studio	http://localhost:1234/v1
vLLM	http://localhost:8000/v1
LiteLLM proxy	http://localhost:4000/v1
llama.cpp server	http://localhost:8080/v1

Security

Use https:// for any server that isn’t on your own computer or network. kunji warns when you add a plain http:// address on the internet, because your key would travel unencrypted.

Remove a server

kunji providers remove lmstudio

Point a built-in provider somewhere else

To send a built-in provider through a proxy, set its base_url:

[providers.openai]
base_url = "https://my-proxy.example.com/v1"