Ollama

// Overview

Ollama is a tool for downloading and running open large language models on your own computer. You install it on macOS, Windows, or Linux, then pull and chat with a model straight from the terminal with a single command.

It is aimed at developers who want to run models locally instead of calling a hosted service. Beyond the CLI, Ollama exposes a REST API on localhost and ships official Python and JavaScript libraries, so you can wire local models into your own apps and scripts.

As a local runtime, Ollama handles model downloads, serving, and the request loop for you, and connects to existing coding tools and agents such as Claude Code, Codex, and Copilot CLI through its launch integrations.

// What it does

• One-line install script for macOS, Windows, and Linux, plus an official Docker image
• Run any model from the library with a single `ollama run` command
• Built-in REST API on http://localhost:11434 for running and managing models
• Official Python (`pip install ollama`) and JavaScript (`npm i ollama`) client libraries
• Launch integrations for coding tools and agents like Claude Code, Codex, and Copilot CLI
• Built on the llama.cpp backend for local model inference

// Getting started

Install Ollama, then pull and chat with a model from the terminal. The same models are reachable over a local REST API and the Python and JavaScript libraries.

curl -fsSL https://ollama.com/install.sh | sh

ollama run gemma4

curl http://localhost:11434/api/chat -d '{
  "model": "gemma4",
  "messages": [{
    "role": "user",
    "content": "Why is the sky blue?"
  }],
  "stream": false
}'

pip install ollama

from ollama import chat

response = chat(model='gemma4', messages=[
  {
    'role': 'user',
    'content': 'Why is the sky blue?',
  },
])
print(response.message.content)

Commands and code are distilled from the project's own documentation — always check the official repo for the latest.

// When to use it

• Run open LLMs locally for privacy or offline work, without sending data to a hosted API
• Add a local model backend to a Python or JavaScript app through the official libraries
• Connect Ollama to coding tools and agents such as Claude Code, Codex, or Copilot CLI
• Prototype and test prompts against different models from the library before committing to a provider

// Version history

Every verified update to Ollama that AI/TLDR tracked, newest first — each links to our coverage and the official changeset.

1. 2026-09-14v0.34.1

   MLX safetensors support in ollama create is no longer experimental. Creating a GGUF model now requires llama.cpp tooling for safetensor conversion and quantization. The release also improves MLX memory handling on Apple Silicon, raises repeat-token detection to 100 tokens, and cuts the /api/tags cold load from 3.1 seconds to 294 milliseconds.

   our coverage →changeset ↗
2. 2026-09-05v0.34.0-rc1

   Ollama models can now be used directly in ChatGPT Desktop, with setup available from the Ollama app on macOS. The release also improves structured output performance on Apple Silicon and adds support for OpenAI-compatible client tool search and response compaction.

   our coverage →changeset ↗

// How Ollama compares

Ollama alongside other open-source local runtimes tools AI/TLDR tracks, ranked by GitHub stars.