Clother supports three local inference backends. No API key is needed for any of them. You must always passDocumentation Index
Fetch the complete documentation index at: https://mintlify.com/jolehuit/clother/llms.txt
Use this file to discover all available pages before exploring further.
--model because local providers have no built-in default model.
| Launcher | Backend | Port |
|---|---|---|
clother-ollama | Ollama | 11434 |
clother-lmstudio | LM Studio | 1234 |
clother-llamacpp | llama.cpp | 8000 |
Ollama
Ollama manages and serves local models with a simple CLI.Install Ollama
Download and install Ollama from ollama.com.
The auth token is set to the literal string
ollama by Clother. You do not need to configure an API key.LM Studio
LM Studio provides a desktop GUI for downloading and running models locally.Install LM Studio
Download and install LM Studio from lmstudio.ai/download.
Load a model
Open LM Studio and download a model from its built-in model browser. Once downloaded, load it.
Start the local server
In LM Studio, navigate to the Local Server tab and start the server on port
1234.The auth token is set to the literal string
lmstudio by Clother. You do not need to configure an API key.llama.cpp
llama.cpp is a high-performance inference engine for GGUF models.Build llama.cpp
Clone and build the project following the instructions at github.com/ggml-org/llama.cpp.
Start the server
model.gguf with the path to your model file. The --jinja flag is required for correct tool-calling support.llama.cpp uses
auth_mode: none. No authentication token is sent.