Continue
Continue is an open source AI coding assistant that plugs into VS Code and JetBrains. It adds chat, inline autocomplete, targeted edits, and agent mode inside the editor you already use, so you can ask questions, refactor code, and run multi-step tasks without switching apps.
You bring your own model provider. Continue supports OpenAI, Anthropic, Google, Mistral, Azure, Amazon Bedrock, Ollama, and many others, and lets you assign different models to different roles like chat, edit, autocomplete, and embeddings. Configuration lives in a YAML file you can version with your repo.
Continue also ships a terminal CLI called cn for AI-assisted workflows outside the IDE. In 2026, Continue announced it was acquired by Cursor. The open source codebase and documentation remain available, and the VS Code and JetBrains extensions still install from their marketplaces.
Agent mode handles multi-step development tasks with tool calling and MCP server support
Chat mode answers questions about your codebase without leaving the editor
Edit mode rewrites selected code sections from natural language instructions
Autocomplete suggests inline completions as you type
Continue CLI (cn) brings the same agent workflows to the terminal
Connect any supported LLM provider or run local models through Ollama
Customize prompts, rules, and model roles through config.yaml
Works inside VS Code and JetBrains instead of forcing a new editor
Wide model provider support with no vendor lock-in
Open source codebase you can inspect and self-host workflows around
Separate model roles for chat, edit, autocomplete, and embeddings
Local model support through Ollama for privacy-sensitive work
Requires manual setup of API keys and config.yaml compared to all-in-one editors
Agent mode depends on models with tool calling support
Acquisition by Cursor leaves long-term roadmap less clear for standalone users
Codebase indexing needs AVX2 and FMA on Linux x64 or those features are disabled
Is Continue free to use?
The Continue VS Code extension, JetBrains plugin, and cn CLI are open source under Apache 2.0. You do not pay Continue for the software itself. Model usage is billed by whichever provider you connect, or you can run free local models with Ollama.
Which IDEs does Continue support?
Continue offers extensions for Visual Studio Code and JetBrains IDEs including IntelliJ, PyCharm, and WebStorm. Install from the VS Code Marketplace or JetBrains plugin repository, then open the Continue sidebar to start.
Can Continue run local models offline?
Yes. Continue supports Ollama and other local model setups. You configure models in config.yaml and can run entirely offline when your local stack is set up. See the Ollama and offline guides in the Continue docs for setup steps.
What happened to Continue after the Cursor acquisition?
Continue.dev announced that Continue was acquired by Cursor. The homepage states the open source project remains available and the team thanked the community for building the codebase. Docs and extension install links are still live.
How do I configure models in Continue?
Edit config.yaml in your workspace or global Continue config directory. Add model entries with provider, API keys via environment variables or .env files, and optional capabilities like tool_use. Reload the config from the agent selector in the sidebar after changes.

