Continue
Open-source AI coding assistant for VS Code and JetBrains.
01 What is Continue?
Continue is a free, open-source AI coding assistant that adds a chat sidebar and inline autocomplete to VS Code and JetBrains. It supports any LLM including local models via Ollama, giving full control over which AI powers your coding experience.
02 Best For
- ▸Open-source AI code assistant
- ▸Local LLM integration
- ▸VS Code and JetBrains users
- ▸Privacy-first AI coding
03 How to Get Started
- 01
Install Continue from the VS Code or JetBrains marketplace.
- 02
Configure your preferred model (Ollama, OpenAI, Anthropic, etc.) in config.json.
- 03
Open the chat sidebar to ask questions or select code and ask for edits.
Pro Tips
- ◆Configure it with your preferred model (including local/open models) — Continue is editor tooling, not a single model.
- ◆Use inline edit (Cmd+I style) for targeted changes and chat mode for broader questions about the codebase.
- ◆Add custom context providers (docs, codebase search) so answers reflect your actual project, not generic knowledge.
- ◆Great pick if you want to self-host or control which model sees your code.
04 At a Glance
Strengths
- +Open-source AI code assistant
- +Local LLM integration
- +VS Code and JetBrains users
- +Privacy-first AI coding
- +Free tier you can start with today
Things to consider
- !Pricing and features change often — confirm current details on the official site.
- !AI output should be reviewed before you rely on it.
Placement reflects our editorial assessment of fit for this use case, not paid placement. Some links may be affiliate links — this never changes our rankings. How we rank →
06 Frequently Asked Questions
What is Continue?
An open-source AI coding assistant for VS Code and JetBrains that lets you plug in any model — cloud or local — for autocomplete and chat.
Is Continue free?
Yes — the extension itself is free and open-source; costs depend on which model/API you connect.
Why choose Continue over Copilot?
Model flexibility and privacy — you can point Continue at local or self-hosted models instead of sending code to a single vendor.