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

# Or on macOS with Homebrew
brew install ollama

# Pull a model
ollama pull llama3.1:8b

# Run interactively
ollama run llama3.1:8b

# List installed models
ollama list

# Serve the API (starts automatically, but can be explicit)
ollama serve

from openai import OpenAI

client = OpenAI(
    base_url="http://localhost:11434/v1",
    api_key="ollama"  # required but ignored
)

response = client.chat.completions.create(
    model="llama3.1:8b",
    messages=[
        {"role": "system", "content": "You are a helpful coding assistant."},
        {"role": "user", "content": "Write a Python function to merge two sorted lists."}
    ],
    temperature=0
)

print(response.choices[0].message.content)

// Switch between local and hosted with an env var
const client = new OpenAI({
  baseURL: process.env.LLM_BASE_URL || "https://api.openai.com/v1",
  apiKey: process.env.LLM_API_KEY || "ollama"
});

# Modelfile for a code review assistant
FROM llama3.1:8b

SYSTEM """You are a senior software engineer reviewing code.
Focus on: bugs, security issues, performance problems, and readability.
Be direct and specific. No fluff."""

PARAMETER temperature 0.1
PARAMETER num_ctx 8192

ollama create code-reviewer -f Modelfile
ollama run code-reviewer

ollama pull deepseek-coder-v2:16b
ollama pull codellama:34b

ollama pull phi3:mini
ollama pull gemma2:9b

# Models I keep loaded
ollama pull llama3.1:8b        # General tasks, chat, brainstorming
ollama pull deepseek-coder-v2:16b  # Code generation and review
ollama pull nomic-embed-text   # Embeddings for local RAG

ollama pull llama3.1:8b         # Default quantization (usually Q4_K_M)
ollama pull llama3.1:8b-q8_0    # 8-bit quantization
ollama pull llama3.1:8b-q4_K_M  # 4-bit quantization (recommended)

# Local RAG pipeline — all on-device
from ollama import Client

client = Client()

# Embed the query
query_embedding = client.embeddings(
    model="nomic-embed-text",
    prompt="How does our authentication flow work?"
)

# Search local vector store (using chromadb locally)
results = collection.query(
    query_embeddings=[query_embedding["embedding"]],
    n_results=5
)

# Generate answer with local model
context = "\n\n".join(results["documents"][0])
response = client.chat(
    model="llama3.1:8b",
    messages=[
        {"role": "system", "content": f"Answer based on this context:\n{context}"},
        {"role": "user", "content": "How does our authentication flow work?"}
    ]
)

#!/bin/bash
# .git/hooks/prepare-commit-msg
# Generate commit message with local model

DIFF=$(git diff --cached)

if [ -z "$DIFF" ]; then
    exit 0
fi

MSG=$(echo "$DIFF" | ollama run llama3.1:8b "Write a concise commit message for this diff. One line, imperative mood, max 72 chars. Output only the message, nothing else.")

echo "$MSG" > "$1"

# ~/.zshrc
alias ai="ollama run llama3.1:8b"
alias aicode="ollama run deepseek-coder-v2:16b"
alias explain="ollama run llama3.1:8b 'Explain this simply:'"