

What is Refact AI?
Refact AI is an AI-powered coding assistant that offers context-aware code completion, in-IDE chat, and a customizable toolbox for software developers working on complex projects. It speeds up coding by providing relevant suggestions based on the entire codebase, while its on-premise deployment option and fine-tuning capabilities make it suitable for companies with strict data security requirements.
What sets Refact AI apart?
Refact AI stands out with its ability to fine-tune models on a company's specific codebase, allowing software engineers to receive suggestions that match their unique coding style and practices. Supporting over 25 programming languages and integrating seamlessly with popular IDEs like VSCode and JetBrains, it's particularly helpful for developers working on large-scale projects with intricate codebases. Refact AI's adaptability makes it a valuable asset for teams looking to speed up their development process while maintaining consistency across diverse programming environments.
Refact AI Use Cases
- Code autocompletion
- In-IDE chat assistance
- Codebase-aware AI
- Custom AI commands
- Self-hosted deployment
Who uses Refact AI?
Features and Benefits
- Refact AI uses Retrieval-Augmented Generation and fine-tuning to provide context-aware code suggestions based on your entire coding environment.
Accurate Code Completion
- Ask questions and interact with the AI directly within your IDE, utilizing the context of your entire codebase for more relevant responses.
Integrated In-IDE Chat
- Access a set of tools for explaining, summarizing, refactoring code, writing documentation, and finding bugs, with the option to create custom commands.
Customizable Toolbox
- Choose from over 20 large language models for chat and code completion, including options like Mistral, Llama3, GPT-4, Code LLama, and StarCoder.
Multiple Language Model Options
- Deploy Refact AI on your own servers or private cloud to maintain control over your data and ensure code privacy.
On-Premise Deployment
Refact AI Pros and Cons
Supports 25+ programming languages
Offers code completion and refactoring
Provides context-aware chat functionality
Can be deployed on-premises for data control
Limited user feedback available
Requires setup and fine-tuning for optimal use
May have a learning curve for new users
Effectiveness may vary depending on codebase
Pricing
Free TrialCode completions powered by Refact 1.6 model
In-IDE Chat powered by GPT-3.5
Toolbox (F1) with in-line code commands
2048 context length for completions
4096 context length for chat
Data training on permissively licensed code
Self-hosting option available
Discord support
Code completions powered by StarCoder2-3B model
More AI models for in-IDE Chat: GPT-4 Turbo, GPT-4o, Claude 3.5 Sonnet
More AI models for Toolbox
x2 context length for completions
x4 context length for chat
Codebase-aware AI for code completions and chat (RAG)
LLM fine-tuning: Train AI models on your organization's codebase and data
Optimized for multiple GPUs with load sharing
Access control for detailed statistics
On-prem or private cloud deployment
Complete code privacy with zero telemetry leaving
Priority support