Complete guide to Tabnine

Getting started

Install the Tabnine extension in your IDE (VS Code, JetBrains, among others) from tabnine.com. Start writing code as usual: you'll see gray suggestions you can accept by pressing Tab or simply ignore by continuing to type. Give Tabnine 1-2 weeks of use so it learns your code patterns and refines its suggestions, and use Tabnine Chat when you need to request refactoring or test generation with context from your open files.

Use cases

Advanced setup

1. Context Engine
   An engine that understands your codebase's full architecture to give suggestions that are more consistent with the project, not just the open file.
2. BYO LLM (bring your own model)
   Lets you connect your own language model instead of relying solely on Tabnine's default models, useful in environments with specific requirements.
3. On-premise mode (Enterprise)
   On the Enterprise plan, all AI inference happens within the company's own infrastructure, with no external calls.

Tips

• Give it time (1-2 weeks) before judging suggestion quality — Tabnine improves noticeably as it learns your code's specific patterns.
• Use Tabnine Chat for tasks that need context from several open files, instead of the inline suggestions meant for quick autocomplete.
• If your team has strict security restrictions on code, evaluate the Enterprise on-premise plan instead of the standard version.
• Combine inline autocomplete for quick tasks with chat for bigger tasks like full refactors.

FAQ