Stay free if you only need limited monthly credits for cloud-routed models, unlimited use of locally-hosted jetbrains models, basic ai assistant features inside supported ides.. Upgrade if you need substantially higher cloud model quotas, more generous junie agent usage, priority access to new models and features, intended for heavy daily users.. Most solo builders can start free.
Why it matters: Requires a JetBrains IDE subscription plus separate AI subscription, making total cost significant ($120-720+ per user annually depending on tier)
Available from: AI Pro
JetBrains AI is supported across the full commercial IDE lineup, including IntelliJ IDEA, PyCharm, WebStorm, GoLand, Rider, PhpStorm, RubyMine, CLion, RustRover, and DataGrip, as well as the Android Studio plugin variant. Both AI Assistant and the Junie agent are available in these environments, though some features roll out to specific IDEs first before becoming generally available.
Yes. JetBrains AI is model-agnostic and lets you route requests to Anthropic Claude, OpenAI GPT, and Google Gemini models, as well as JetBrains-hosted local models for offline use. You can switch models per chat or task, which is useful for balancing latency, cost, and answer quality for different kinds of work.
JetBrains states that customer code is not used to train third-party models, and enterprise customers can opt out of all data collection. Local models keep code entirely on-device, and cloud-routed requests can be configured with data-residency preferences. Admins also have access to audit logs and centralized policy controls.
AI Assistant is the inline and chat-based helper that handles completion, code generation, refactoring suggestions, and conversational Q&A. Junie is JetBrains' agentic coding companion that takes a higher-level task description, plans a sequence of edits, executes them across multiple files, runs tests, and iterates autonomously while you supervise from the IDE.
Yes, JetBrains AI ships with comprehensive MCP support, allowing teams to connect custom MCP servers for internal documentation, issue trackers, databases, deployment systems, and proprietary tooling. This lets the assistant reason about organization-specific context that is not present in the public training data of the underlying models.
Start with the free plan — upgrade when you need more.
Get Started Free →Still not sure? Read our full verdict →
Last verified March 2026