ollama

Overview

Ollama is a tool for running open-source large language models locally. It supports a variety of models including Kimi, GLM, MiniMax, DeepSeek, gpt-oss, Qwen, and Gemma. With a simple CLI and REST API, you can download, run, and manage models on your own machine.

Key Features

• Supports a wide range of open models including Kimi, GLM, MiniMax, DeepSeek, gpt-oss, Qwen, and Gemma.
• Provides a simple CLI to run and chat with models.
• Offers a REST API for programmatic access.
• Includes official libraries for Python and JavaScript.
• Available on macOS, Windows, Linux, and via Docker.
• Can launch integrations with coding tools like Claude Code, Codex, Copilot, and more.

Use Cases

• Coding: Use Ollama to run local models for code assistance with integrations like Claude Code and Codex.
• AI assistant: Turn Ollama into a personal AI assistant across messaging platforms using OpenClaw.
• Chat with models: Run and chat with models like Gemma 4 directly from the terminal.
• REST API: Integrate Ollama into applications via its REST API for chat and model management.

Getting Started

• Install Ollama on macOS, Windows, or Linux using the provided install scripts, or download the installer. For Docker, use the official image. After installation, run 'ollama' to get started, then use 'ollama run <model>' to chat with a model. See the quickstart guide for more

Deployment & Requirements

• Ollama can be deployed on macOS, Windows, Linux, or via Docker. It requires a compatible system with sufficient resources to run models locally.

Before You Adopt

• License: MIT. Review its terms before using, modifying, or distributing the project.