Skip to main content

Setup Ollama with Llama 3

Use your own local AI model for CV rewriting and quality assessment. Free, private, and runs entirely on your computer.

Video Tutorials

Prefer to watch a video? Here are helpful resources and search terms to find step-by-step video tutorials:

Find Video Tutorials on YouTube

Search YouTube for step-by-step video guides. Recommended search terms:

  • "ollama install tutorial"
  • "llama 3 setup ollama"
  • "ollama llama 3.2 installation"
  • "install ollama [macOS/Windows/Linux]"
  • "ollama beginner tutorial"
Search YouTube Tutorials →

Official Ollama Resources

The official Ollama website with guides, examples, and documentation.

Visit ollama.com →

Tip: Many video tutorials cover the same steps as this written guide, but watching someone do it can be helpful if you're a visual learner. Look for recent videos (2024-2025) to ensure they cover the latest Ollama and Llama 3 versions. Popular channels include tutorials from developers, tech educators, and AI enthusiasts.

What is "Local AI"?

"Local AI" means running an AI model directly on your own computer, instead of sending your data to a cloud service. Think of it like having a personal assistant that works entirely on your device - no internet connection needed after setup, and your information stays private.

How it works:

  1. You install Ollama (a free application) on your computer
  2. You download an AI model (like Llama 3) to your computer
  3. The AI runs entirely on your device - your CV data never leaves your computer
  4. You can use it anytime, even without internet

Why Use Local AI Instead of Cloud Services?

Here are the key advantages of running AI on your own computer:

  • Free: No API costs or subscription fees - use it as much as you want
  • Private: Your CV data never leaves your computer - complete privacy
  • Fast: No internet required after initial setup - works offline
  • Unlimited: No rate limits or usage restrictions - use it as often as you need
1

Install Ollama

Ollama is a free, open-source tool that runs AI models locally on your computer.

Download and Install Ollama

  1. Visit ollama.com/download in your web browser
  2. Click the "Download for macOS" button
  3. Open your Downloads folder and double-click the downloaded file (it will be named something like Ollama-darwin.dmg)
  4. A window will open showing the Ollama app icon. Drag the Ollama icon to your Applications folder
  5. Open your Applications folder (press Cmd+Shift+A or click Applications in Finder)
  6. Double-click Ollama to open it. You may see a security warning - if so:
    • Go to System Settings (or System Preferences on older macOS)
    • Click "Privacy & Security"
    • Click "Open Anyway" next to the Ollama message
    • Confirm you want to open Ollama

Open Terminal (Where You'll Type Commands)

  1. Press Cmd + Space to open Spotlight Search
  2. Type "Terminal" and press Enter
  3. A black or white window will open - this is where you'll type commands
  4. You should see a prompt that looks like: yourname@yourmac ~ %

Verify Ollama is Installed

  1. In the Terminal window, type: ollama --version
  2. Press Enter
  3. If you see a version number (like "ollama version is 0.x.x"), Ollama is installed correctly!
  4. If you see "command not found", make sure Ollama is running (check your Applications or the menu bar)

Which Model Should You Use?

Different models offer different balances of quality, speed, and resource requirements. Choose the one that best fits your computer and needs.

Llama 3.2 (3B)

Best for: Most users, balanced performance

Recommended

Pros

  • Fast responses (10-30 seconds)
  • Low memory usage (~4GB RAM)
  • Small download (~2GB)
  • Good quality for CV tasks
  • Works on most computers

Cons

  • Less nuanced than larger models
  • May miss subtle context
  • Shorter responses

System Requirements: 8GB+ RAM, ~4GB disk space

Command: ollama pull llama3.2

Llama 3.1 (8B)

Best for: Better quality, more powerful computers

Pros

  • Higher quality responses
  • Better understanding of context
  • More nuanced suggestions
  • Better for complex CVs

Cons

  • Slower responses (30-60 seconds)
  • High memory usage (~16GB RAM)
  • Large download (~9GB)
  • May struggle on older computers

System Requirements: 16GB+ RAM, ~9GB disk space

Command: ollama pull llama3.1:8b

Llama 3.3

Best for: Latest features and improvements

Latest

Pros

  • Latest improvements and fixes
  • Better performance than 3.2
  • Multiple size variants available
  • Active development

Cons

  • Size varies by variant (2-10GB)
  • Less tested than 3.2
  • May have occasional bugs

System Requirements: Varies by variant (8-16GB+ RAM)

Command: ollama pull llama3.3

Mistral 7B

Best for: Alternative high-quality option

Pros

  • Excellent quality
  • Good balance of speed/quality
  • Well-optimized
  • Strong performance

Cons

  • Requires 12GB+ RAM
  • Larger download (~4GB)
  • Slower than 3.2

System Requirements: 12GB+ RAM, ~4GB disk space

Command: ollama pull mistral

Phi-3 (Microsoft)

Best for: Very fast, low-resource computers

Pros

  • Very fast responses (5-15 seconds)
  • Very low memory (~2GB RAM)
  • Tiny download (~2GB)
  • Works on older computers
  • Good for basic tasks

Cons

  • Lower quality than larger models
  • Less nuanced responses
  • May struggle with complex CVs

System Requirements: 4GB+ RAM, ~2GB disk space

Command: ollama pull phi3

Recommendation

Start with Llama 3.2 - it offers the best balance of quality, speed, and resource usage for CV tasks. If you have a powerful computer (16GB+ RAM) and want higher quality, try Llama 3.1 (8B). If you have limited resources, Phi-3 is a good lightweight option.

2

Download Your Chosen Model

Once you've chosen a model from the comparison above, download it using the commands below.

Check Available Disk Space First

Before downloading, make sure you have enough free disk space:

  1. Click the Apple menu (top left) → "About This Mac"
  2. Click "Storage" tab
  3. Check your available space - you'll need at least 5-10GB free depending on the model

Recommended: Llama 3.2 (3B parameters)

Best balance of quality and speed. Works well on most computers with 8GB+ RAM.

Disk Space Required: ~2GB download, ~4GB total after installation

  1. Open Terminal (press Cmd + Space, type "Terminal", press Enter)
  2. In the Terminal window, type exactly: ollama pull llama3.2
  3. Press Enter
  4. You'll see download progress. This will take several minutes depending on your internet speed
  5. When you see "pulling complete", the model is ready!

ollama pull llama3.2

Alternative: Llama 3.1 (8B parameters)

Higher quality results, but requires more RAM (16GB+ recommended).

Disk Space Required: ~4.7GB download, ~9GB total after installation

  1. In Terminal, type: ollama pull llama3.1:8b
  2. Press Enter and wait for download to complete

ollama pull llama3.1:8b

Latest: Llama 3.3

Most recent version with latest improvements.

Disk Space Required: ~2-5GB download (varies by variant), ~4-10GB total after installation

  1. In Terminal, type: ollama pull llama3.3
  2. Press Enter and wait for download to complete

ollama pull llama3.3

Important Notes:

  • The first download will take several minutes depending on your internet speed
  • Model files are large (2-10GB depending on the model) - make sure you have enough disk space before starting
  • Subsequent uses will be instant as the model is stored locally on your computer
  • If you run out of space during download, you can delete the partial download and try again later
3

Verify Your Setup is Working

Make sure Ollama is running and your model is ready to use.

1. Make Sure Ollama is Running

Ollama needs to be running on your computer for the AI features to work.

  • macOS/Windows: Check if the Ollama app is open. If not, open it from your Applications folder
  • Linux: Ollama should start automatically, but if not, run: ollama serve

2. Verify Your Model is Installed

Check that your downloaded model is available:

macOS/Linux: Open Terminal and run:

ollama list

Windows: Open Command Prompt and run:

ollama list

You should see your downloaded model (e.g., llama3.2) in the list.

3. Test Ollama API Connection

Verify that Ollama's API is accessible:

macOS/Linux: In Terminal, run:

curl http://localhost:11434/api/tags

Windows: In Command Prompt, run:

curl http://localhost:11434/api/tags

This should return a JSON list of your installed models. If you see an error, make sure Ollama is running.

Configure Your Connection

Once Ollama is installed and your model is downloaded, you need to configure Simple CV Builder to use your local Ollama installation.

You'll need to be logged in to configure your AI settings. After logging in, go to the AI Settings page to enter your Ollama connection details.

4

Configure Connection in Simple CV Builder

Enter Your Ollama Settings

Now that Ollama is installed and your model is downloaded, configure Simple CV Builder to connect to it:

You'll need to be logged in to configure your AI settings. After logging in:

  1. Go to the AI Settings page
  2. Select "Local Ollama" as your AI Service
  3. Enter your Ollama Base URL (usually http://localhost:11434)
  4. Enter your model name (e.g., llama3.2)
  5. Click "Test Connection" to verify it works
  6. Click "Save Settings"

Try the AI Features

Once configured, you can test the AI features in Simple CV Builder:

CV Quality Assessment

Get AI-powered feedback on your CV quality with scores and recommendations.

You'll need to be logged in to use this feature.

AI CV Rewriting

Generate job-specific CV variants automatically from job descriptions.

You'll need to be logged in to use this feature.

What to Expect

  • The first request may take 30-60 seconds as the model loads into memory
  • Subsequent requests will be faster (10-30 seconds)
  • All processing happens on your computer - your data never leaves your machine
  • If you see an error, check that Ollama is running and the model is installed

Troubleshooting

Ollama not found

Make sure Ollama is running. On macOS/Windows, check if the Ollama app is open. On Linux, start it with: ollama serve

Connection refused

Verify Ollama is running on port 11434. Check with: curl http://localhost:11434/api/tags

Model not found

Make sure you've downloaded the model. Run ollama pull llama3.2 (or your chosen model).

Slow performance

Try a smaller model like llama3.2 instead of larger ones. Also ensure you have enough RAM (8GB+ recommended).

Out of memory errors

Close other applications to free up RAM, or use a smaller model. Llama 3.2 (3B) requires less memory than Llama 3.1 (8B).

You're All Set!

Now you can use AI-powered CV features completely free and privately on your own computer.

Get Started → Learn More About AI Features →

CV Prompt Best Practices — learn how to write effective prompts for AI CV rewriting.