Refact AI

What is Refact AI

Refact AI is an open-source AI coding assistant designed to enhance coding efficiency with fast code completion, powerful code improvement tools, and chat functionalities. It supports various language models like GPT-4, Code Llama, StarCoder, and WizardCoder, and can be integrated into IDEs such as JetBrains and VS Code. Users can choose between cloud-hosted or self-hosted versions, optimizing privacy and customization based on their needs.

Key Features of Refact AI

- Cloud-Hosted AI Coding Assistant: Provides a robust and scalable coding assistant hosted in the cloud, with per-user statistics for tracking productivity and usage.

- Integration with Popular IDEs: Compatible with JetBrains and VS Code, allowing seamless integration into your existing coding environment.

- Best-in-Class LLMs: Utilize leading language models (LLMs) for coding, including GPT-4, Code Llama, StarCoder, WizardCoder, and more.

- Customizable Code LLM Selection: Users can experiment and choose the language models that best fit their specific coding requirements.

- Context-Aware Code Suggestions: Employs a fill-in-the-middle technique to understand the context of your codebase, providing accurate and relevant suggestions.

- Self-Hosted Options: Offers self-hosted versions for users who prefer not to share their code externally, providing flexible privacy settings.

- Automatic Code Completion: Enables automatic completion of code lines as you type, enhancing productivity and reducing coding time.

- Pre-Built Prompts for Code Refactoring: Provides pre-built prompts to refactor, analyze, and fix bugs in your code, making coding more efficient.

- Interactive Code Chat: Allows users to ask any code-related question in an interactive chat feature, providing real-time assistance and solutions.

Pricing

Free Plan

• Cost: $0/month
• Best for: Personal use and hobby projects
• Features:
  • Code completion powered by advanced Refact gen2 model
  • GPT-3.5 Powered Chat
  • GPT-3.5 Powered AI Toolbox functions
  • 4096 context tokens in chat
  • Cloud or self-hosting options
  • Discord support
  • Subject to Fair Use Policy

Professional Plan

• Best for: More complex coding tasks and professional use
• Features:
  • Everything included in the Free plan plus:
  • Code completion powered by advanced Refact gen2 model
  • GPT-4 Powered AI Toolbox functions
  • GPT-4 Powered Chat
  • 16384 context tokens in chat
  • Data training only on permissively licensed code
  • Repo-level context-awareness for code completions and chat (RAG)

Enterprise Plan

• Custom pricing
• Best for: Teams needing a secure and customizable AI code platform
• Features:
  • Everything included in the Professional plan plus:
  • Fine-tuning on your code
  • Optimized for several GPUs with load sharing
  • Access control
  • On-prem deployment
  • Complete code privacy without 3rd-party dependencies
  • Priority support
  • Contact required for custom pricing details