llmserve
A terminal UI that finds locally installed LLM inference backends and model files, then lets a user launch and monitor local model servers without writing config files or CLI flags.
llmserve is a terminal UI for serving local LLM models. It auto-detects installed inference engines such as llama-server, KoboldCpp, LocalAI, and MLX, discovers model files across common locations such as an LM Studio directory or the Hugging Face cache, and lets a user launch a server for any combination with live log output.
The interface has three panels: a source tree of model locations on the left, a searchable and sortable model table in the center, and running server cards with live backend logs on the right. Keybindings support panel focus cycling, resizing, navigation, adding custom model directories, and filtering by name or format.
It installs via a shell script, Homebrew, or Cargo, and is written in Rust. A companion project helps determine which models fit a given machine's hardware before llmserve runs them.