TRAE

Visit TRAE

TraeCode integrates seamlessly into your workflow, collaborating with you to maximize performance and efficiency.

No ratings yet|6

From $0

Code Vertical

Alternatives

AgentMail

Gives AI agents inboxes so they can send and act on email.

Grok Bot

Always-on AI teammates that get their own cloud computer, sign into your tools, and finish multi-step work on their own.

Is Agentic

Score how ready a website is for AI agents, then get evidence and recommendations to improve it.

Make This Better

Triage feedback. Ship improvements. Build a better product

On this page

What is TRAE?

TRAE (styled Trae) is an AI-native IDE and coding agent published by SPRING PTE, a Singapore-based entity under ByteDance — the Chinese technology company behind TikTok/Douyin, founded in 2012 in Beijing by Zhang Yiming. Trae runs as a desktop IDE (macOS and Windows) built around a familiar code-editor layout, plus a browser-based Cloud IDE for anyone who doesn't want to install anything locally, and it supports the Model Context Protocol (MCP) for connecting external tools and APIs, along with configurable custom agents built around specific tools, skills, and logic.

On March 31, 2026, Trae launched SOLO, a standalone agent app (desktop and web) that no longer requires the IDE plugin, in two modes: Code (a focused, chat-driven coding loop where the agent writes, runs, and iterates on code) and MTC — "More Than Coding" — aimed at a broader set of roles (product managers, data scientists, marketing, and operations) who use SOLO conversationally to create documents, analyze data, generate reports, build presentations, and run competitive research rather than write code. SOLO Desktop opens local files as a project workspace, while SOLO Web runs entirely in the cloud.

Core Features

AI-Native IDE

A full desktop code editor (macOS and Windows) with AI-assisted autocomplete, in-editor Q&A, and agent-based programming for multi-step coding tasks.

Cloud IDE

A browser-based version of the IDE for development without installing anything locally.

SOLO — Code Mode

A standalone, chat-driven coding agent that writes, runs, and iterates on code autonomously based on a described task, without requiring the full IDE.

SOLO — MTC Mode

"More Than Coding" mode: a conversational agent for non-coding work like document creation, data analysis, report generation, presentations, and competitive research.

Model Context Protocol (MCP) Support

Connects to external tools and APIs via MCP, and supports configuring custom agents with specific tools, skills, and logic.

Token-Based Usage Billing

AI work draws from a Basic Usage and Bonus Usage balance, with separate fast and slow request pools for premium models, rather than a simple fixed monthly request count.

How to Use TRAE

Download the Trae desktop IDE for macOS or Windows, or open the browser-based Cloud IDE, and sign in.
Work inside the editor as you would with any code-focused IDE, using Trae's AI features for code completion, in-editor Q&A, and agent-based programming that can carry out multi-step coding tasks.
For a standalone, non-IDE agent experience, use SOLO instead: choose Code mode for a chat-driven coding loop, or MTC mode for broader, non-coding work like documents, data analysis, reports, and presentations.
Connect external tools and APIs via MCP, or configure a custom agent with a specific toolset, skills, and logic for a repeatable task.
Track your monthly usage against your plan's Basic Usage and Bonus Usage balance — Trae bills AI work as token-based usage rather than a simple fixed request count, with separate "fast" and "slow" request pools for premium models.
Upgrade from the Free plan to Lite, Pro, Pro+, or Ultra as usage grows, or contact BytePlus for enterprise deployment.

Use Cases

Everyday software development inside a full IDE, with AI-assisted autocomplete, in-editor Q&A, and agent-based programming for multi-step coding tasks.
A chat-driven, hands-off coding loop via SOLO's Code mode, where the agent writes, runs, and iterates on code based on a described task.
Non-coding knowledge work handled conversationally via SOLO's MTC mode — assembling documents, analyzing datasets, generating reports, building presentations, or running competitive research — for roles like product management, data science, and marketing.
Cloud-based development without installing anything locally, using Trae's browser-based Cloud IDE or SOLO Web.
Connecting external tools, APIs, and data sources via MCP, or building a custom agent scoped to a specific tool set and task.
Enterprise deployment through BytePlus for teams that need it managed outside the consumer product.

Pros & Cons

Pros

One product spans both a full AI-native IDE for hands-on coding and a standalone SOLO agent for hands-off, chat-driven work — including a non-coding MTC mode aimed at roles beyond engineering
Available as a native desktop app, a browser-based Cloud IDE, and a standalone SOLO web app, so there's a path in that doesn't require a local install
The Free plan is genuinely usable rather than a stripped-down trial: 5,000 autocompletions and both fast and slow premium-model requests per month at no cost
Supports MCP and configurable custom agents, so it can be extended with external tools and repeatable task-specific logic rather than being limited to what ships out of the box

Cons

Trae shifted from a fully free product at launch to a five-tier paid structure (Free, Lite $3, Pro $10, Pro+ $30, Ultra $100) introduced around February 2026, so older reviews describing it as "completely free" are now out of date
As a Singapore-registered ByteDance product with a Chinese parent company, organizations with vendor-risk or data-residency policies restricting Chinese-linked software should evaluate Trae's data handling before adopting it — independent write-ups of the product have previously raised questions about its data collection
SOLO launched in beta on March 31, 2026 and rolled out region by region (starting in select markets, with the US and Canada following shortly after), so availability and feature completeness can vary by region and account
The token-based Basic/Bonus Usage billing model, with separate fast and slow request pools, is more complex to reason about than a flat per-month request count, making it harder to predict costs in advance

Pricing

Free

5,000 AI autocompletions per month
10 fast and 50 slow premium model requests per month
Limited Basic Usage
2 concurrent cloud tasks

Lite

$3/month

More Basic Usage than Free
Entry-level paid tier for light usage

Pro

$10/month (after a 7-day free trial)

Included SOLO mode
Unlimited autocomplete
Fast queue priority
Up to 10 concurrent cloud tasks

Pro+

$30/month

Higher usage allowance than Pro
Built for heavier agentic workflows

Ultra

$100/month

Highest usage tier
Multi-task cloud execution for context-heavy sessions

Enterprise

Custom (via BytePlus)

Managed deployment outside the consumer product
Contact BytePlus for details