Continue — Open-Source AI Code Assistant for IDEs

Core Features

Chat with Code Context

Select code and ask questions with full codebase awareness:

Ctrl+L → Opens chat sidebar
Select code → "Explain this function"
@ mention files → "@auth.ts how does this relate to the login flow?"

Autocomplete

Tab-complete with any model:

{
  "tabAutocompleteModel": {
    "title": "Codestral",
    "provider": "mistral",
    "model": "codestral-latest"
  }
}

Inline Editing

Select code and describe changes:

Ctrl+I → "Add error handling and input validation"
Continue generates the edit, you accept or reject

Context Providers

Pull context from multiple sources:

Model Flexibility

Use any model from any provider:

{
  "models": [
    {"title": "Claude", "provider": "anthropic", "model": "claude-sonnet-4-20250514"},
    {"title": "GPT-4", "provider": "openai", "model": "gpt-4o"},
    {"title": "Local Llama", "provider": "ollama", "model": "llama3.1"}
  ]
}

Custom Slash Commands

{
  "customCommands": [
    {
      "name": "test",
      "prompt": "Write comprehensive unit tests for the selected code. Use the existing test framework."
    }
  ]
}

Key Stats

• 22,000+ GitHub stars
• VS Code + JetBrains support
• 20+ model providers
• Fully customizable config
• Active community with 500+ contributors

FAQ

Q: What is Continue? A: Continue is an open-source AI code assistant for VS Code and JetBrains that lets you connect any LLM for chat, autocomplete, and inline editing with full customization.

Q: Is Continue free? A: Yes, fully open-source under Apache 2.0 license. You provide your own model API keys.

Q: How is Continue different from GitHub Copilot? A: Continue is open-source, supports any LLM provider (not just OpenAI), and is fully customizable — you control the models, context, and commands.