ByteDance agentic IDE with autonomous project building, multi-model switching, and free frontier model access.
No compliance attestations on file. Confirm directly with the vendor before procurement.
| Tier | Price | Includes |
|---|---|---|
Free | Free | Limited basic usage, 5,000 autocompletions/month, 2 concurrent cloud tasks, standard queue, limited TRAE SOLO mode access |
Lite | $3/seat/mo | — |
Pro | $10/seat/mo | — |
Pro+ | $90 basic usage + bonus, up to 15 concurrent cloud tasks, fast queue. Pricing varies by region; not yet publicly listed in USD on the pricing page. | — |
Ultra | $400 basic usage + bonus, up to 20 concurrent cloud tasks, model early access. Pricing varies by region; not yet publicly listed in USD on the pricing page. | — |
What it does Trae is an agentic IDE from ByteDance, forked from VS Code. Provides chat, multi-model switching, autocomplete, and TRAE SOLO — an autonomous project-building mode that breaks specs into tasks and executes them across files. Available as desktop and web clients.
Who it's for Individual developers wanting a free or low-cost alternative to Cursor with frontier model access bundled in. Hobbyists, students, and developers working on non-sensitive personal projects. Not appropriate for commercial code, client work, or any context with confidentiality obligations.
How platform engineers use it In practice, Trae has limited adoption inside enterprise platform teams because of the documented telemetry concerns. Where it is used, it's typically as a sandbox for trying agentic IDE patterns on disposable repos. The IDE itself supports the standard VS Code workflow with AI chat, autocomplete, and SOLO autonomous mode. Concurrent cloud-task quotas (2 on Free/Lite, 10 on Pro, 15 on Pro+, 20 on Ultra) determine how much parallel agent work you can run.
Strengths
Limitations
AI maturity AI-native in product design — TRAE SOLO and the multi-model switching are real agentic features, not skins on chat. The model access is competitive. The maturity issue is operational: the product appears built for fast user acquisition over data minimization, which materially limits its appropriate use cases.
