Ollama

The Ollama node lets you run powerful AI models directly on your computer instead of using cloud services. This means complete privacy (your data never leaves your machine), no ongoing costs, and the ability to work offline.

Perfect for sensitive data processing, cost-conscious projects, or when you want full control over your AI processing.

How it works

Ollama runs as a local server on your computer, hosting AI models that your workflows can connect to. Instead of sending data to external APIs, everything processes locally with complete privacy.

graph LR
  Workflow[Your Workflow] --> Ollama[Local Ollama Server]
  Ollama --> Model[AI Model]
  Model --> Response[Private Response]
  style Ollama fill:#6d28d9,stroke:#fff,color:#fff

Setup guide

1. Install Ollama: Download from ollama.com and install on your computer.

2. Download AI Models: Run ollama pull llama2:7b to download your first model.

3. Start the Server: Run ollama serve to start the local AI server.

4. Connect Workflows: Use http://localhost:11434 as your Ollama URL in workflows.

Practical example: Private document analysis

Let’s set up local AI processing for analyzing sensitive documents.

Option 1: Basic Setup

• Server URL: http://localhost:11434 (Default local address)
• Model: llama2:7b (Good balance of speed and quality)
• Creativity: Low (0.3) for consistent answers.

Option 2: Creative Writing

• Model: mistral:7b (Known for good writing)
• Creativity: High (0.8) for more varied output.
• Length: Up to 1000 tokens (longer responses).

Option 3: Code Analysis

• Model: codellama:7b (Specialized for code)
• Creativity: Very low (0.1) for precision.
• Length: Up to 800 tokens.

Why choose local AI

Local AI (Ollama)	Cloud AI Services
Complete privacy - data never leaves your machine	Data sent to external servers
No ongoing costs after setup	Pay per API call
Works offline	Requires internet connection
No rate limits	API rate limits apply
Full control over models	Limited model choices

Popular models to try

Model	Best For	RAM Needed	Speed
llama2:7b	General tasks, good balance	8GB	Fast
mistral:7b	Fast responses, efficient	8GB	Very Fast
codellama:7b	Code analysis and generation	8GB	Fast
llama2:13b	Higher quality responses	16GB	Slower

Start with llama2:7b

The llama2:7b model provides excellent quality for most use cases while being reasonably fast and not requiring too much RAM.

System requirements

Minimum requirements

• RAM: 8GB (for 7b models)
• Storage: 4GB free space
• CPU: Modern processor (2018+)

Recommended setup

• RAM: 16GB+ (for larger models)
• Storage: 10GB+ free space
• GPU: Optional, for faster processing

Real-world examples

Private document summarization

Process confidential documents without external services:

Setup: llama2:7b with temperature 0.3
Use case: Summarize legal documents, medical records, financial reports
Benefit: Complete privacy, no data leaves your computer

Offline content generation

Create content without internet connection:

Setup: mistral:7b with temperature 0.7
Use case: Write emails, articles, creative content
Benefit: Works anywhere, no connectivity required

Code review and analysis

Analyze proprietary code safely:

Setup: codellama:7b with temperature 0.2
Use case: Code review, bug detection, documentation
Benefit: No code sent to external services

Troubleshooting

• “Connection failed” errors: Make sure Ollama is running with ollama serve and check the URL is correct.
• “Model not found” errors: Download the model first using ollama pull model-name.
• Slow responses: Try a smaller model (7b instead of 13b) or check if your system has enough RAM.
• Out of memory errors: Close other applications or use a smaller model that fits your available RAM.