Ollama Complete Guide: The Easiest Way to Run LLMs Locally

Want to run large models locally, but llama.cpp requires compilation, vLLM needs configuration, and LocalAI demands dependency installation. Is there something that works right out of the box? Yes, Ollama.

💡 What You Will Learn

Want to run large models locally, but llama.cpp requires compilation, vLLM needs configuration, and LocalAI demands dependency installation. Is there something that works right out of the box? Yes, Ol

📜 Table of Contents

• OllamaWhat Is
• Installation
• macOS / Linux
• Windows

OllamaWhat Is

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

# 
ollama run qwen2.5

Installation

macOS / Linux

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

Windows

ollama --version

# Auto/Automatic Qwen2.57B
ollama run qwen2.5

# Parameter
ollama run qwen2.5:14b  # 14B
ollama run deepseek-r1:7b  # DeepSeek R1
ollama run llama3.2:3b  # 

|:----|:----|:---:|:---------:|:--------| | Qwen2.5-7B | ollama run qwen2.5 | 4.5GB ||| | DeepSeek-R1-7B | ollama run deepseek-r1:7b | 4.5GB ||| | Qwen2.5-14B | ollama run qwen2.5:14b | 8.5GB ||| | Llama 3.2-3B | ollama run llama3.2:3b | 2.2GB ||| | Gemma 3-27B | ollama run gemma3:27b | 16GB ||| | CodeGemma | ollama run codegemma | 5GB |||

#  Ollama  :11434
curl http://localhost:11434/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{
    "model": "qwen2.5",
    "messages": [{"role": "user", "content": ""}]
  }'

# Modelfile
FROM qwen2.5
SYSTEM "Python"

ollama create my-python-assistant -f Modelfile
ollama run my-python-assistant