GatwAI

JetBrains Junie

JetBrains Junie across IDE and terminal: plans, model choice, cloud credits, local execution and limitations on complex engineering tasks.

Provider sources checked

What does JetBrains Junie do?

Junie is JetBrains' development agent for planned coding work. The current interface describes terminal use, IDE connections and execution through GitHub or GitLab. Requirements and implementation can be broken into reviewable plans, and ongoing tasks can be steered. Custom guidelines and skills aim to reuse project routines. Its distinction is combining agent work with development tools and selectable models. For JetBrains users this can fit existing workflows more closely than a separate application builder. Evaluate a maintenance task in your own project: which files changed, what was actually tested and how much review remained? A provider benchmark does not answer those questions.
Editorial research · 2026-11-14How we research →

Who is this tool for?

  • Developers using JetBrains tools or terminal workflows.
  • Teams with reusable coding conventions and controlled model selection.

Strengths

  • Planning, implementation and review can be connected structurally.
  • Model selection and bring-your-own-key enable different cost models.
  • IDE tools and persistent project rules support recurring work.

Limitations and requirements

  • Local models need suitable hardware and may produce different results.
  • Cloud credits and your own model API costs must be considered separately.
  • Broad changes still need review even with IDE inspections.

Who is behind the product?

Company
JetBrains s.r.o.
Headquarters
Not separately established by the reviewed sources.
Company founded
Not separately established by the reviewed sources.
Product launch
Introduced in January 2025 with an early-access waiting list.

The official introduction initially described Junie as an agent inside JetBrains IDEs. Today's offering also includes a CLI and local models. The introduction date is distinguished from current capabilities and company history.

Suggested ways to use it

Suggested workflows or verification steps. These are not hands-on tests; for unverified products they are not a recommendation to use the service.

Suggested: restructure an existing function

  1. Specify behavior and the permitted change scope.
  2. Review Junie's plan before permitting edits.
  3. Assess inspections, regression tests and the diff together.

Goal: A structured refactoring proposal with traceable validation.

Suggested: recurring maintenance convention

  1. Describe project conventions and test commands as guidelines.
  2. Run a small maintenance task using those guidelines.
  3. Correct deviations and reuse only proven rules.

Goal: A reproducible workflow instead of informal chat instructions.

Pricing, allowances and data

What should you check before choosing a plan?

The current page lists Junie Lite as a free entry point and bring-your-own-key with separate model charges. AI Pro includes ten AI credits per 30 days and AI Ultimate 35, with top-ups available. Displayed monthly prices depend on annual/monthly selection and customer type and should be confirmed at checkout. Local execution avoids cloud consumption but requires suitable hardware.

What should you check about data and permissions?

JetBrains states in its knowledge base that inputs and outputs are not used for training without consent; third-party providers have corresponding restrictions but may perform limited safety processing. The current data notice distinguishes telemetry and voluntary collection. Junie Local is a separate local-execution route; an ordinary cloud session does not thereby become local.

External reviews and available evidence

No GatwAI hands-on test. This assessment uses current provider sources, not an independent original test of today's version.

Reported positives

  • A reviewable plan and IDE inspections provide concrete checkpoints.

Reported cautions

  • Benchmarks and provider examples do not establish results for your project. Suggested workflows have not been executed by GatwAI.

Which alternative fits which task?

Cursor →

For a dedicated integrated agent development environment.

Frequently asked questions

Does Junie always run locally?

No. Cloud models, your own API keys and local models are different operating routes.

Does one AI credit equal a completed task?

Not generally. Consumption depends on model and work; credits cannot be equated with completed assignments.