How to Install Ollama on Ubuntu Server

Ollama makes running large language models on Linux as simple as running a Docker container. One command installs everything — no CUDA setup, no Python dependencies.

Step 1 — Install Ollama

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

Step 2 — Pull and Run a Model

# Pull Llama 3 (latest Meta model)
ollama pull llama3

# Run interactively
ollama run llama3

# Pull Mistral (faster, lighter)
ollama pull mistral
ollama run mistral

Step 3 — List Available Models

ollama list
# Shows: NAME, ID, SIZE, MODIFIED

# See all available models
# https://ollama.com/library

Step 4 — Use the REST API

# Ollama runs API on port 11434 by default
curl http://localhost:11434/api/generate   -d '{"model":"llama3","prompt":"What is nginx?","stream":false}'

# Chat API (OpenAI compatible)
curl http://localhost:11434/v1/chat/completions   -H "Content-Type: application/json"   -d '{"model":"llama3","messages":[{"role":"user","content":"Explain Docker in one paragraph"}]}'

Step 5 — Configure Ollama Service

# Ollama installs as systemd service automatically
sudo systemctl status ollama
sudo systemctl restart ollama

# Change listen address (default: localhost only)
sudo nano /etc/systemd/system/ollama.service

[Service]
Environment="OLLAMA_HOST=0.0.0.0:11434"
Environment="OLLAMA_NUM_PARALLEL=2"
Environment="OLLAMA_MAX_LOADED_MODELS=2"

sudo systemctl daemon-reload
sudo systemctl restart ollama

Step 6 — Secure with Nginx

server {
    listen 443 ssl;
    server_name ai.yourdomain.com;

    auth_basic "AI Server";
    auth_basic_user_file /etc/nginx/.htpasswd;

    location / {
        proxy_pass http://127.0.0.1:11434;
        proxy_set_header Host $host;
        proxy_read_timeout 300s;
    }
}

Step 7 — Manage Models

# Delete a model
ollama rm llama2

# Copy/rename
ollama cp llama3 my-custom-llama

# Show model info
ollama show llama3

Conclusion

Ollama is the fastest way to get AI running on a Linux server. Our team installs and configures Ollama and AI server environments for production use.