TraeCode
The AI programming IDE under ByteDance's TRAE offers, according to the official statement, a dual development mode of IDE and SOLO. It has a built-in Coding Agent that can take over tasks and schedule tools, and supports custom agents and MCP.
Updated: 2026-10-26
What is TraeCode?
TraeCode is an AI development IDE under TRAE, officially positioned as "Your Super AI Development Engineer". It is operated by Beijing Yinli Dangdiao Technology Co., Ltd. (ByteDance). It is an independent product line from its sibling TraeWork (an AI office platform): TraeWork covers office scenarios such as PPT, data analysis and in-depth research, while TraeCode focuses on writing code. The self-description on the official website's homepage is "An AI IDE product under TRAE, with intelligent productivity as its core".
Who is it for?
- Developers who want an "AI-led development" experience and let Agents automatically break down and advance tasks.
- Senior engineers who prefer to control the pace themselves and only use AI as a completion and conversational assistant (using IDE mode).
- Teams that need to consolidate team specifications into reusable rules (TRAE Rules) and shared intelligent agents.
- Individual developers who want a unified development environment across macOS, Windows, and Linux.
Key features
6Dual Development Modes of IDE and SOLO
The official states that the IDE mode retains the original development process and offers stronger control; the SOLO mode is AI-led, with a built-in dedicated Coding Agent that can understand goals, plan tasks, schedule tools, and automatically advance development through all stages.
Cue Intelligent Prediction and Completion
The official states that it achieves precise continuation based on an optimized model, intelligently perceives the warehouse context and editing trajectory, automatically predicts the next modification point, allows jumping to apply suggestions by pressing Tab, and provides multi-line editing suggestions and diff comparison.
Custom Agents & Agent Marketplace
You can freely define tools, skills, and task logic to build exclusive agents; it supports independently creating Agents and sharing them to the marketplace for combined use like plugins.
MCP and External Tool Integration
Empowers agents to call external resources through the Model Context Protocol.
Context Understanding & TRAE Rules
The official states that in addition to in-IDE information such as code repositories and terminal details, it supports internet searches and document set uploads, and allows the configuration of custom AI workflow rules (TRAE Rules).
Privacy and Security
The official statement emphasizes a "local-first" approach and "minimal data collection." Code files are saved on local devices by default. Any data temporarily uploaded for index generation is deleted in plaintext after computation. It also offers privacy mode, an ignore function, and regional deployment.
How to use TraeCode
- First, decide which mode to use: If you're starting a task from scratch and want AI to take the lead, choose SOLO mode.
- if you have an existing project and need precise control over every change, select IDE mode.
- You can switch between the two at any time.
- Free version available: The official pricing page states that the free tier includes 500 credits per month and that "all features are available for free", supporting 2 concurrent cloud tasks.
- new users get an additional 4500 credits upon downloading and logging in.
- It is recommended to first use the free credits to complete a real project.
- The points for paid tiers are consumed in a unified unit based on usage: Pro ¥99/month, Pro+ ¥239/month, Ultra ¥699/month (the official note states that there is a discounted price for the first month).
- The main differences lie in the number of points, the concurrency of cloud tasks, and priority access to new models.
- most tiers indicate model discounts (e.g., 75% off for Seed-2.1-Turbo, 48%/47% off for GLM, and 50% off for DeepSeek), with the actual discounts subject to the purchase page.
- Pay attention to the benefits of the two product lines: The official pricing page clearly states that "Lite membership is only available for TraeWork", so confirm which product line's quota you are purchasing before buying.
- It is recommended to configure TRAE Rules and Privacy Mode first before getting started: Proper rule configuration can significantly reduce invalid changes.
- when dealing with sensitive repositories, use Privacy Mode first to limit the scope of data usage.
Traffic & popularity
Data month Sep 2026
TraeCode receives about 1.5M visits per month and ranks #35,738 worldwide.
Monthly visits
12 monthsTraffic sources
- Direct55.6%
- Search28.1%
- Display ads7.6%
- Referrals5.5%
- Social3.2%
Top countries
- China89.8%
- Hong Kong, China3.5%
- United States2.2%
- Singapore1.4%
- Japan0.7%
Estimated from third-party traffic data.
In these collections
Similar tools
CodeBuddyTencent Cloud has launched an AI full-process intelligent programming tool, which the official says is available in three forms: IDE, editor plug-ins and CLI. Based on conversational large models such as Hunyuan, it covers demand analysis, design draft to code conversion, code completion and one-click deployment.
Google AntigravityGoogle has launched an agent development platform that offers desktop, command-line, IDE extensions and SDK forms, enabling parallel management of multiple local agents; the personal version is free, with additional Pro and Ultra subscription tiers.
Lyzo AILyzo is an AI-native standalone IDE. Officially, it features Claude-powered completion and conversation, full codebase indexing, cross-session persistent memory and autonomous engineering agents. It also integrates zero-configuration CI/CD, Docker and automatic deployment, supporting more than 40 programming languages.
ReplitReplit is a cloud-based IDE and AI application building platform. In the browser, use Replit Agent to generate and deploy full-stack applications conversationally, with built-in authentication, databases, hosting, and integrations, suitable for going from prototype to production.
ClineOpen-source AI programming agent, officially said to adopt the Plan-then-Act mode to modify code across files and execute bash commands in editors and terminals, supporting MCP extensions and multi-Agent collaboration.
KiroAmazon has launched an AI integrated development environment, with specification-driven as the main line: writing prompts into executable specifications, verifying the correctness of code, using parallel agents to work in large code bases, and its official positioning is to move from assisting in writing code to agent engineering.