What it is

Continue is an open-source coding agent designed to assist developers directly within their IDE or terminal. It operates as a CLI tool, VS Code extension, and JetBrains plugin, enabling natural-language interaction with codebases to generate, edit, and explain code. The project lives in the AI-assisted developer tools ecosystem and is built with TypeScript under the Apache-2.0 license.

The project solves the problem of fragmented AI coding assistance by offering a unified, local-first agent interface that integrates with multiple development environments. It removes reliance on proprietary tooling by providing a self-hostable, transparent alternative for code generation and refactoring tasks.

Key capabilities

• Execute natural-language code edits via slash commands in the IDE or terminal
• Support for custom configuration via config.json to define models, tools, and prompts
• Run as a standalone CLI using @continuedev/cli for terminal-based workflows
• Integrate with VS Code via extension marketplace and OpenVSX registry
• Provide JetBrains integration via plugin, though the CLI is recommended over the plugin
• Remove anonymous telemetry and authentication in the final 2.0.0 release for privacy and simplicity
• Support extensibility through custom tools and model configurations

Who uses it and how

• Developers use the CLI to run code generation and refactoring tasks directly in terminal sessions
• VS Code users install the extension to trigger agent actions via Ctrl+L or / commands in the editor
• Teams self-host Continue to avoid external API dependencies and maintain control over model usage
• Contributors extend the agent by modifying the TypeScript codebase or adding custom tools
• JetBrains users deploy the plugin for in-IDE assistance, though migration to the CLI is recommended

Getting started

Install the CLI via npm install -g @continuedev/cli, or install the VS Code extension from the marketplace or OpenVSX. The JetBrains plugin is available via GitHub releases. Configuration begins with generating a default config.json file.

When to use it — and when not to

Continue is suitable for developers seeking a privacy-focused, self-hostable coding agent with CLI and IDE integration. It replaces paid tools by offering similar core capabilities without subscription fees, but requires local setup of models and dependencies. Users must manage model access (e.g., via OpenAI, Anthropic, or local LLMs) and may need to configure storage or network access depending on use case. The project is no longer actively maintained and is read-only, limiting future feature development or bug fixes.