Skip to content

LOCAL-SETUP.md

Latest commit

 

History

History
160 lines (120 loc) · 3.43 KB

LOCAL-SETUP.md

File metadata and controls

160 lines (120 loc) · 3.43 KB

DeepSeek CLI - Local Setup Guide

This guide will help you set up DeepSeek CLI to run completely locally using Ollama, providing free and private AI assistance.

🎯 Benefits of Local Setup

  • Free: No API costs or usage limits
  • Private: Your code never leaves your machine
  • Fast: No network latency for requests
  • Offline: Works without internet connection
  • Customizable: Full control over models and settings

📋 Prerequisites

  1. Node.js 18+: Download here
  2. Ollama: Download here
  3. Sufficient RAM:
    • 2GB+ for deepseek-coder:1.3b
    • 8GB+ for deepseek-coder:6.7b (recommended)
    • 32GB+ for deepseek-coder:33b

🚀 Quick Installation

Option 1: Automated Script (macOS/Linux)

curl -fsSL https://raw.githubusercontent.com/holasoymalva/deepseek-cli/main/install-local.sh | bash

Option 2: Manual Installation

  1. Install Ollama:

    # macOS
brew install ollama

# Linux
curl -fsSL https://ollama.ai/install.sh | sh

# Windows: Download from https://ollama.ai

  2. Install DeepSeek CLI:

    npm install -g run-deepseek-cli

  3. Start Ollama:

    ollama serve

  4. Install DeepSeek Model:

    # Choose one based on your system:
ollama pull deepseek-coder:1.3b    # Lightweight (1GB)
ollama pull deepseek-coder:6.7b    # Recommended (4GB)
ollama pull deepseek-coder:33b     # Most capable (19GB)

  5. Test Installation:

    deepseek chat "Write a hello world function in Python"

🔧 Configuration

Create a .env file in your project or home directory:

# Local mode (default)
DEEPSEEK_USE_LOCAL=true
DEEPSEEK_MODEL=deepseek-coder:6.7b
OLLAMA_HOST=http://localhost:11434

📖 Usage Examples

Interactive Mode

deepseek

Single Prompt

deepseek chat "Explain how async/await works in JavaScript"

Force Local Mode

deepseek --local --model deepseek-coder:1.3b

Setup Helper

deepseek setup

🛠️ Troubleshooting

Ollama Not Running

# Start Ollama service
ollama serve

# Or run in background
nohup ollama serve > /dev/null 2>&1 &

Model Not Found

# List installed models
ollama list

# Install missing model
ollama pull deepseek-coder:6.7b

Connection Issues

# Check if Ollama is accessible
curl http://localhost:11434/api/tags

# Use custom host
deepseek --ollama-host http://192.168.1.100:11434

Memory Issues

  • Use smaller model: deepseek-coder:1.3b
  • Close other applications
  • Increase system swap space

🔄 Switching Between Local and Cloud

Use Local (Default)

export DEEPSEEK_USE_LOCAL=true
deepseek

Use Cloud API

export DEEPSEEK_USE_LOCAL=false
export DEEPSEEK_API_KEY="your-api-key"
deepseek

📊 Model Comparison

Model Size RAM Needed Speed Quality Best For
deepseek-coder:1.3b 1GB 2GB Fast Good Quick completions
deepseek-coder:6.7b 4GB 8GB Medium Better General coding
deepseek-coder:33b 19GB 32GB Slow Best Complex analysis

🎉 You're Ready!

Your DeepSeek CLI is now configured to run locally. Enjoy free, private, and powerful AI coding assistance!

For more advanced features and configurations, check the main README.md.