Skip to the content.

Ollama

Information

Ollama is a local LLM runtime for downloading, serving, and running AI models on your own machine. It is commonly used for private local AI workflows, development experiments, API-based integrations, and testing models without relying on a hosted cloud inference provider.

If you use Ollama as a local runtime, these model families are worth knowing because they are commonly used for general reasoning, coding, math, multilingual work, and research-oriented experimentation.

Practical Ollama note

Model availability in Ollama depends on what has been packaged or imported for Ollama usage. Some of these families may be available directly from the Ollama library, while others may require checking the latest model names, tags, quantizations, or community packaging status before pulling them locally.

Ollama Setup Guide (Local LLM Runtime)

This guide covers installing, running, and managing models with Ollama for local AI workflows.


1. Install Ollama

Linux / macOS:

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

Verify installation:

ollama --version

▶️ 2. Start Ollama Server

ollama serve

By default, Ollama runs at:

http://localhost:11434

3. Download Models

Use ollama pull to download models.

# General reasoning
ollama pull llama3:8b
ollama pull mistral

# Coding models
ollama pull deepseek-coder:6.7b
ollama pull codellama:13b

# Larger / more powerful (requires strong GPU)
ollama pull llama3:70b
ollama pull deepseek-coder:33b

4. Run a Model

Start an interactive session:

ollama run llama3:8b

Example prompt:

Explain how a REST API works

5. Use Ollama as an API

Ollama exposes an OpenAI-compatible API.

Example request:

curl http://localhost:11434/api/generate -d '{
  "model": "llama3:8b",
  "prompt": "Write a Python function to reverse a string"
}'

6. List Installed Models

ollama list

7. Remove a Model

ollama rm llama3:8b

8. Update a Model

ollama pull llama3:8b

(Re-pulling updates the model)


9. Environment Variables (Optional)

export OLLAMA_HOST=0.0.0.0
export OLLAMA_PORT=11434

10. Recommended Models by Use Case

🟢 General Use

Coding

High Performance


Notes

See also