Skip to content

How to Install Bolt.new Locally (bolt.diy) — and How It Compares to Bolt.new

In the rapidly evolving landscape of AI-powered development tools, Bolt has emerged as a groundbreaking solution that combines cutting-edge AI models with StackBlitz’s web container technology. This comprehensive guide explores the differences between the official Bolt.new platform and its open-source local installation, helping developers make an informed decision about which version best suits their needs.

How to install Bolt.new locally

Bolt.new itself runs in your browser at bolt.new, so there is nothing to install. “Installing Bolt” usually means running bolt.diy, the community-maintained open-source version you can run on your own machine with your own API keys or local models. bolt.diy is a separate community project, not an official StackBlitz product.

Requirements

  • Node.js 22.12.0 or later
  • pnpm 9.15.9 or later (npm install -g pnpm)
  • An API key for at least one provider (Anthropic, OpenAI, Gemini, and others), or a local model through Ollama or LM Studio

Install and run

git clone https://github.com/stackblitz-labs/bolt.diy.git
cd bolt.diy
pnpm install
pnpm run dev

Then open http://localhost:5173 in your browser.

Add your API keys

Open Settings (the gear icon) → Providers and paste your key. Alternatively, copy .env.example to .env.local and set variables such as ANTHROPIC_API_KEY.

Docker option

cp .env.example .env.local
docker compose --profile development up

Desktop app

Prebuilt Windows, macOS, and Linux builds are available on the bolt.diy releases page.

Push your project to GitHub

Whichever version you use, the free Bolt to GitHub extension pushes Bolt.new projects to a GitHub repository in one click.

Understanding Bolt

Bolt.new represents a significant advancement in development environments, integrating AI models with full control over the development ecosystem. This integration enables developers to run complete NodeJS servers, manage npm packages, and deploy to production—all directly within their browser. The platform’s distinguishing feature lies in how it grants AI models comprehensive control over the entire development environment, from file system operations to package management.

Local Installation vs Official Platform

Installation Requirements

The setup steps for the current open-source version, bolt.diy, are in How to install Bolt.new locally above.

Common Features

Both versions share several powerful core capabilities:

  • An intuitive chat interface for AI interaction
  • Code editor with syntax highlighting
  • Live preview functionality
  • File management system
  • Terminal access
  • Code download options

Unique Features of Official Bolt.new

The official platform offers several advantages not present in the local installation:

  1. Enhanced File Management
    • Context-specific right-click menu
    • File targeting system for limiting AI changes
    • File locking capability to prevent modifications
  2. Advanced Code Interaction
    • Interactive code selection with “Ask Bolt” functionality
    • Improved full-screen preview mode
    • Token usage tracking
  3. Deployment Options
    • Direct Netlify integration
    • Same-URL deployment capabilities
    • Streamlined project sharing

Advantages of Local Installation

The local version provides unique benefits:

  1. LLM Flexibility
    • Support for multiple AI providers
    • Direct control over API usage
    • Custom model selection options
  2. Interface Customization
    • Chat panel toggle
    • Code preview panel controls
    • Enhanced terminal integration
  3. Development Control
    • Full control over the development environment
    • Direct file system access
    • Custom configuration options

Practical Comparison: Building a Git Simulator

To demonstrate the capabilities of both versions, we implemented a git simulator application. This practical example revealed several key differences:

  1. Code Generation
    • Official Version: More refined initial output but occasionally requires multiple attempts to resolve issues
    • Local Version: More straightforward generation process with fewer iterations needed
  2. Performance
    • Official Version: Benefits from ongoing optimizations and updates
    • Local Version: Depends on local resources but offers more consistent performance
  3. Development Experience
    • Official Version: More polished UI with integrated tools
    • Local Version: Greater flexibility but requires more manual intervention

Making the Right Choice

The decision between using Bolt.new or its local installation depends on your specific needs:

Choose Bolt.new if you need:

  • Quick prototyping capabilities
  • Seamless deployment integration
  • Access to the latest features
  • Minimal setup requirements

Opt for Local Installation if you require:

  • Custom LLM integration
  • Full environment control
  • Offline development capabilities
  • Specialized configuration options

Future Considerations

While the open-source version hasn’t received updates for over a month, the community continues to develop forks with enhanced features. This creates opportunities for specialized implementations tailored to specific development needs.

Conclusion

Both versions of Bolt offer compelling advantages for different use cases. The official Bolt.new platform excels in providing a seamless, feature-rich experience ideal for rapid development and deployment. The local installation, while requiring more setup, offers greater control and customization options. Developers can even leverage both versions, using GitHub integration to maintain project synchronization across platforms.

For developers seeking to make the most of these tools, understanding these differences is crucial for selecting the right version for their specific needs. As the platform continues to evolve, both versions will likely see further improvements and innovations, making them increasingly valuable assets in the modern development toolkit.

Stuck on an AI-built app? If your Bolt / Lovable / Cursor / Claude Code project is fragile, won’t deploy, or has auth or database issues, I can help with AI app rescue. Request a Code Rescue Audit →

Leave a Reply

Your email address will not be published. Required fields are marked *