Skip to main content

Using DeepSeek with Ollama and Brave

Introduction

This is a quick-start guide for leveraging Ollama's local AI capabilities with Brave browser while maintaining control over your data and privacy.

Target Audience

  • AI Newcomers: Those taking their first steps beyond ChatGPT into the broader AI ecosystem
  • DeepSeek Users: Current users of chat.deepseek.com and/or DeepSeek mobile apps
  • Privacy-Conscious Users: Those interested in DeepSeek models but seeking more control over their data
  • Cost-conscious users going deeper: Those interested in running AI models locally on their own hardware

Prerequisites

  1. Install Ollama

  2. Install Brave Browser

Setting Up Brave with Ollama

  1. Install the DeepSeek Model
    Open Terminal, Command Prompt (cmd), bash, or PowerShell and run:

    ollama pull deepseek-r1:14b
  2. Open Brave Settings

    • Click the menu icon (☰) in the top-right corner
    • Select "Settings"
    • Click on "Leo"
      Brave Leo Settings
  3. Configure Model

    • Scroll to the bottom and click "Add new model" with the following settings:
      Label: DeepSeekR1_14b
      Model request name: deepseek-r1:14b
      Server Endpoint: http://localhost:11434/v1/chat/completions
      Brave Ollama Model Configuration
    • Click "Save"

Verifying Your Setup

  1. Test the Connection

    • Click the Leo icon in Brave
    • Select your configured DeepSeek model
    • Try a simple test prompt like "Hello, how are you?"
  2. Troubleshooting

    • If Leo can't connect, verify Ollama is running:
      ollama list
    • Check that the model is installed:
      ollama list
    • Verify the server endpoint is accessible:
      curl http://localhost:11434/v1/chat/completions

Privacy Benefits

By running Ollama locally:

  • All AI interactions stay on your machine
  • No data leaves your control
  • Complete privacy for sensitive queries
  • No dependency on external services
  • Full control over model usage and data

Next Steps

  • Experiment with different models available through Ollama
  • Explore Leo's features in Brave
  • Try more complex prompts and use cases
  • Consider setting up multiple models for different purposes

Additional Resources


Last updated: January 30, 2025