The Command Line Interface (CLI) lets you run `.litertlm` models on Linux,
macOS, Windows, Android, or Raspberry Pi.

> [!TIP]
> **Tip:** For performance benchmarks and a full list of supported models, check the LiteRT-LM [Overview](https://developers.google.com/edge/litert-lm/overview).

## Overview of CLI Capabilities

The LiteRT-LM CLI is a powerful tool for interacting with local LLMs.
It supports:

- **Quick Installation** : Run on-demand with `uvx` or install permanently using `uv` or `pip`. See [Installation](https://developers.google.com/edge/litert-lm/cli/installation).
- **Interactive Chat** : Run models in an interactive chat session. See [Usage](https://developers.google.com/edge/litert-lm/cli/usage).
- **Configuration** : Customize global defaults and per-model settings in `config.json`. See [Configuration](https://developers.google.com/edge/litert-lm/cli/configuration).
- **Advanced Capabilities** : Optimize and extend your models with [Multi-Token Prediction](https://developers.google.com/edge/litert-lm/cli/usage#mtp), [Multi-modality](https://developers.google.com/edge/litert-lm/cli/usage#multi-modality) attachments, and [Function Calling](https://developers.google.com/edge/litert-lm/cli/usage#function-calling).
- **OpenAI-Compatible Server** : Start a local server that emulates OpenAI endpoints. See [OpenAI-Compatible Server](https://developers.google.com/edge/litert-lm/cli/openai_server).

To get started, proceed to the [Installation](https://developers.google.com/edge/litert-lm/cli/installation) guide.