彙整社群經驗的在地化AI程式設計指南

Navigation Menu

Search code, repositories, users, issues, pull requests...

Provide feedback

We read every piece of feedback, and take your input very seriously.

Saved searches

Use saved searches to filter your results more quickly

To see all available qualifiers, see our documentation.

local-ai-coding-guide

License

Uh oh!

There was an error while loading. Please reload this page.

murataslan1/local-ai-coding-guide

Folders and files

Latest commit

History

Repository files navigation

🦙 Local AI Coding Guide

Run GPT-4 class AI coding assistants 100% locally. No API costs. No cloud. Total privacy.

Complete guide with agentic workflows, prompt engineering, runner comparison, and real-world examples

⚡ Quick Links:

🚀 Quick Start · 🤖 Agentic Coding · 🔀 Runners · 🛡️ Guardrails · 🎯 Prompts · 🗣️ Community · ⚠️ Gotchas

📋 Table of Contents

🚀 Getting Started

🔧 Infrastructure

🤖 Advanced Workflows (NEW)

⚠️ Troubleshooting

🛠️ Tools & Configs

🎯 Why Local AI?

2026 Reality: Qwen2.5-Coder-32B scores 92.7% on HumanEval, matching GPT-4o. The switch is no longer a compromise—it's an upgrade.

The Bandwidth Formula

🚀 Quick Start

Step 1: Install Ollama

Step 2: Download Coding Model

Step 3: Test It

Step 4: Install Continue.dev (VS Code)

Done! You now have a local Copilot alternative.

🔀 Runner Comparison

vLLM is 19x faster than Ollama under concurrent load (Red Hat benchmarks).

Quick Decision

📖 Full Runner Comparison Guide →

🤖 Agentic Coding (NEW!)

Reddit's #1 requested feature: "Show me a real workflow, not just setup."

The Bug Fix Workflow (Aider + Ollama)

Example Session

Continue.dev Agent Mode

📖 Full Agentic Coding Guide →

🛡️ Guardrails & Coding Plans

Prevent local models from hallucinating and breaking your code.

Strategy 1: TDD as Feedback Loop

Strategy 2: Plan Before Code

Strategy 3: Scope Limiting

📖 Full Guardrails Guide →

🎯 Prompt Engineering

Local models need better prompts than GPT-4.

The CO-STAR Framework

Identity Reinforcement

System Prompt Template

📖 Full Prompt Engineering Guide →

📊 Model Comparison

Quantization Guidance

Warning: Don't go below Q4 for coding. Logic breaks at low precision.

💻 Hardware Requirements

The Speed Formula

Recommendations

🔧 IDE Integration

Continue.dev (Recommended)

Cursor (Local Mode)

Aider (Terminal)

🖥️ Alternative Tools

📖 Full Alternative Tools Guide →

🔄 Real-World Workflows

Workflow 1: Debug React Component

Workflow 2: Add API Endpoint (TDD)

Workflow 3: Refactor Legacy Code

📖 Full Workflows Guide →

⚠️ Gotchas

Top 5 Mistakes

Context Window Exhaustion

📖 Full Gotchas Guide →

⚡ Optimization Guide

Keep Model in Memory

Increase Context Window

💰 Cost Analysis

Insight: If you already have a gaming PC, local AI is essentially free.

📈 Star History

📚 Resources

🤝 Contributing

We welcome contributions! Help us keep this guide updated.

💝 Support

⭐ Star this repo if it helped you!

Made with ❤️ by Murat Aslan

Last updated: January 2026

About

local-ai-coding-guide

Topics

Resources

License

Contributing

Uh oh!

There was an error while loading. Please reload this page.

Stars

Watchers

Forks

Releases

Packages

  0

Languages

Footer

Footer navigation