TRAE

TRAE is an AI-powered code editor designed to function as a 10x AI Engineer, capable of independently building software solutions. It combines a traditional IDE experience with an autonomous coding agent called SOLO that can handle the entire development lifecycle from requirements to deployment. TRAE understands your vision, plans workflows, executes with precision, and deploys production-ready code.

• TRAE SOLO Mode provides a fully autonomous coding agent that drives the entire development process from start to finish, handling requirements definition, implementation, database integration, and deployment while you stay visually in control.

• Multi-Agent Framework offers built-in agents with specialized expertise out of the box, plus the ability to create custom agents tailored to specific workflows. Agents can plan, execute, and act as sub-agents to tackle complex challenges together.

• Model Context Protocol (MCP) Integration enables agents to access external resources on demand, with a built-in preview tab that turns the browser into a development tool for interacting with elements, reading console logs, and debugging in real-time.

• CUE Autocomplete predicts your next edit with a single Tab keystroke, understanding your intent and anticipating your next move with smart suggestions across multiple lines at once.

• Deep Context Understanding allows TRAE to comprehend your development context through code repositories, online searches, and shared documents, with customizable rules to tailor AI behavior to your workflow.

• Privacy-First Architecture stores codebase files locally with minimal data collection, offering Privacy Mode and ignore functions to control how data is used, with encrypted transmission and regional deployment options.

• Seamless VS Code Migration provides a smooth transition from VS Code with familiar interface patterns and extension compatibility.

To get started, download TRAE from the official website for macOS or Windows. The IDE mode brings agentic capabilities into your traditional coding workflow, while SOLO mode unleashes the fully autonomous coding agent. You can build custom agents, integrate external tools via MCP, and leverage the open agent ecosystem to share and discover community-built agents.