TRAE

What's Trae?

Trae is an AI programming assistant (AI native IDE) launched by ByteDance, which realizes efficient collaboration between human and AI through dialogic interaction. Users can @ Intelligent Body and provide context to let AI complete the requirements.Trae supports customized Intelligent Body, which can flexibly define tools, skills and task logic according to the requirements, and build an exclusive AI team.Trae IDE supports the seamless integration of a variety of external tools, and calls external resources through the MCP protocol to expand the capability boundary.

Trae's main functions

• Intelligent Code Generation and Optimization : Code snippets can be generated through natural language, supporting code completion, optimization and refactoring to improve programming efficiency.
• AI-driven interaction patterns It provides two modes: Chat mode, which supports answering code questions and suggesting code updates, and Builder mode, which generates complete code projects based on requirements.
• Native Chinese Support : Support Chinese from the bottom of the design, the interface language comprehensive Chinese culture, suitable for Chinese developers.
• Integration of mainstream AI models The built-in Claude 4, GPT-4o and other powerful AI models, the domestic version of the beanbag model and DeepSeek model, free to use, to help quickly generate high-quality code.
• Convenient project preview and debugging : Provide Webview function, you can preview web pages directly in the IDE, which is convenient for front-end development.
• Flexible contextual referencing : Support for referencing code blocks, files, folders, or entire projects in AI conversations for precise interactions.
• Multimodal development support : Uploading design drawings or prototype drawings can directly generate code, but also supports screenshots to mark the modification requirements, lowering the threshold of development for non-technical users.

Trae official website address

Trae domestic version of the official website address:https://www.trae.com.cn/

Trae International official website address: https://www.trae.ai/

How to use Trae

• download and install To install Trae, visit Trae's official website, click "Download" to enter the download page, select the appropriate version according to your computer's operating system, run the installer, you can select the theme, language, and choose to import the configuration from VS Code or Cursor.
• Installation of command line tools (optional) : Click the "Install trae commands" button to complete the authorization process, after which you can use the trae command to quickly recall Trae, or use the trae my-react-app command opens the specified project in Trae.
• Login account : When you start Trae for the first time, the domestic version supports cell phone number and Nuggets account login; the international version supports GitHub and Google account login.
• Open an existing project or clone a Git repository : Once you've logged in, you can choose to open an existing project or clone a Git repository to start developing with Trae.
• Create new projects with AI Assistant If you are creating a new project from 0 to 1, you can use Trae's AI assistant. For example, if you switch to Builder mode and enter the project requirements in the dialog box, such as "Please create a Snake web game with React", you will be prompted to specify a directory as the project directory, and then the initialization commands will be executed automatically.
• Experience Smart Code Generation AI will provide real-time code suggestions to realize code auto-completion when you enter code in the code editing area; you can also trigger function-level code completion by typing Chinese comments, such as typing "Calculate User Integral" and pressing Tab key, AI will complete the related function.
• Utilizing the intelligent question and answer function Chat with the AI assistant at any time during the coding process. Right-click on the selected code and select "Ask Trae", or use the shortcut key "Ctrl+Shift+A" to call for Q&A, and ask the AI questions such as "Why is this code reporting an error? "The AI will automatically read the selected code and related files and give you a solution to fix the problem.
• Experience multimodal interaction Support image to code, you can upload design drawings or sketches, AI will directly parse and generate front-end code; you can also upload error screenshots or design sketches, AI can understand the content of the picture and generate the corresponding code.
• Installing command line tools : After clicking the "Install trae commands" button and completing the authorization process, you can use Trae related commands in the terminal, so that you can quickly complete Trae operations in the terminal.
• Customized Intelligentsia The AI team is a team of AI intelligences that flexibly defines the tools, skills, and task logic of the intelligences according to their own needs, so that the AIs can serve them better.
• Explore the plug-in market : Find the Plug-in Marketplace in the Smart Navigation panel on the left and install various plug-ins to enrich and expand Trae's functionality.

Trae's core strengths

• AI Native Design Trae is designed from the ground up with AI as a core component. Unlike traditional IDEs that introduce AI functionality through smart plug-ins, Trae integrates AI with the development process in a deeper way, providing stronger context-awareness and the ability to handle a wider range of development scenarios, and delivering better results in key development tasks such as code auto-completion and bug fixing.
• Intelligent code generation function : It supports generating code snippets through natural language descriptions, and also provides code optimization suggestions based on the contextual environment, helping developers save time and write higher quality code.
• Strong contextual understanding It not only understands information within the IDE, such as code repositories and endpoint information, but also supports the understanding of external information, such as networked searches and uploaded document collections, and configures the AI work rules to allow the AI to perform customized tasks on-demand.
• Intelligent auto-completion : With powerful contextual analytics, code fragments can be predicted and continued in real-time, quickly extending unfinished code and improving coding efficiency several times over.
• multimodal interaction AI supports image to code, uploading design drawings or sketches, and AI will directly parse and generate front-end code, which makes code generation more intuitive and convenient, especially for UI interface development and other scenarios, which can greatly improve development efficiency.
• Rich AI model support : Built-in powerful AI models such as Claude-3.5-Sonnet, GPT-4o, Beanbag 1.5 pro, DeepSeek R1/V3, and more provide Trae with powerful intelligence support, and their use is completely free for 2025.
• Migration Friendly : Support for migrating configurations from tools such as VS Code and Cursor makes it easy for developers to migrate previous settings to Trae without having to reconfigure the entire development environment, reducing the extra workload associated with changing tools.
• Plug-in ecology is rich : Support for installing plug-ins from the VS Code marketplace, with rich extensibility, developers can install a variety of plug-ins to extend the functionality of Trae according to their own needs and usage habits.
• Native Chinese Support : The interface language is fully Chinese, and the variable names of the domestic version have been changed to full Chinese, which is more in line with the language habits of Chinese developers and reduces the difficulty of use caused by the language barrier.

Trae's application scenarios

Trae can be used in many scenarios: in the development field, it helps Web, game, data processing, API and tool development, such as generating code according to the description, previewing and debugging, and providing optimization suggestions, etc.; in learning and teaching, it generates examples and explanations for beginners, and helps teachers prepare for classes; in office collaboration, it facilitates team sharing of documents and real-time feedback; and in creative creation, it assists designers to categorize materials, communicate, and manage versions to improve efficiency in all aspects. In office collaboration, it helps designers to categorize materials, communicate, collaborate, and manage versions, which improves the efficiency of each link in an all-round way.