Trae

Trae (The Real AI Engineer) is ByteDance's free AI-native code editor, launched in January 2025 as a direct challenger to Cursor and Windsurf. Built on VS Code and distributed by ByteDance's Singapore subsidiary Spring (SG) Pte, Trae targets developers outside China with an aggressive free-first strategy: no credit card required, and full access to Claude 4.6 Sonnet, GPT-4o, Gemini 2.5 Pro, and DeepSeek R1 on the free plan. It reached 1.6 million+ monthly active users within its first year, making it one of the fastest-growing AI IDEs globally — particularly strong among developers in Asia and price-sensitive global markets.

Trae's core agentic features include Builder Mode (multi-file code generation that plans, edits, and verifies changes across a codebase from a natural language prompt) and SOLO Mode (fully autonomous project scaffolding that handles requirements analysis, code generation, terminal commands, browser testing, and deployment as a running agent). SOLO launched with Trae 2.0 and became available as a standalone desktop and web app on March 31, 2026, independent of the IDE. A Cloud IDE provides browser-based access with no local installation required. MCP support extends Trae to external tools and databases.

The primary concern with Trae is privacy: security researchers confirmed in 2025 that it collects detailed telemetry including file paths, hardware identifiers, and behavioral metrics even when the telemetry toggle is set to off. ByteDance retains personal data for 5 years after account closure with no opt-out option. Teams working on proprietary codebases should carefully evaluate this risk. Trae offers a token-based pricing model introduced in February 2026, with plans ranging from free to $100/month (Ultra). As the most feature-rich free AI IDE in 2026, Trae is an excellent choice for individual developers and students who can accept the telemetry trade-off for access to frontier models at zero cost.

Trae uses a token/credit-based pricing model introduced in February 2026. The Free plan includes 5,000 autocompletions per month, 10 fast + 50 slow premium model requests, and 1,000 advanced requests — all at no cost, no credit card required.

Lite costs $3/month with increased token allocations. Pro costs $10/month with 600 fast premium model requests per month, no queue waiting, and priority access. Pro+ is $30/month, and Ultra is $100/month for power users and heavy agentic workloads.

All plans include access to Claude 4.6 Sonnet, GPT-4o, GPT-4.1, Gemini 2.5 Pro, DeepSeek R1, and ByteDance's Doubao models. SOLO Mode and Builder Mode are available on the free plan within monthly quota limits. The Cloud IDE and Trae Work platform are included with all plans.