Ollama Setup Guide

Step 1: Install Ollama

macOS:

# Download from ollama.com or use Homebrew
brew install ollama

# Start Ollama
ollama serve

Windows:

# Download installer from ollama.com
# Run installer
# Ollama starts automatically

Linux:

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

# Start Ollama
ollama serve

Step 2: Download Models

Popular Models (2026):

# Llama 3.2 (Meta) - Best all-around
ollama pull llama3.2

# Llama 3.2 3B - Lightweight, fast
ollama pull llama3.2:3b

# Mistral - Great for reasoning
ollama pull mistral

# Gemma 2 (Google) - Creative tasks
ollama pull gemma2

# Phi-3 (Microsoft) - Code specialist
ollama pull phi3

# Nomic Embed - For RAG/vector search
ollama pull nomic-embed-text

Check Downloaded Models:

ollama list

Step 3: Test Ollama

Chat Mode:

# Start interactive chat
ollama run llama3.2

# Type your message, press Enter
# Type /bye to exit

Single Query:

# One-off query
ollama run llama3.2 "What is AI orchestration?"

Step 4: Use with LangChain

from langchain_ollama import ChatOllama
from langchain_core.prompts import ChatPromptTemplate
from langchain_core.output_parsers import StrOutputParser

# Initialize local model
llm = ChatOllama(
    model="llama3.2",
    base_url="http://localhost:11434"
)

# Create prompt
prompt = ChatPromptTemplate.from_messages([
    ("system", "You are a helpful assistant."),
    ("user", "{input}")
])

# Create chain
chain = prompt | llm | StrOutputParser()

# Run
response = chain.invoke({"input": "Explain AI orchestration"})
print(response)

🔧 Troubleshooting