Continue

Continue is the open-source coding agent for your IDE and terminal. It brings AI chat, autocomplete, and autonomous agent workflows into VS Code and JetBrains IDEs, with the freedom to connect any model you like.

Overview

Continue is an open-source project on GitHub with around 36,000 stars and daily activity through 2026. It started as an AI code assistant for VS Code and JetBrains and has grown into a full agent platform: a sidebar chat that understands your codebase, next-edit autocomplete, an agent mode that can plan and apply multi-step changes, and a CLI for terminal and CI workflows. Through the Continue Hub you can assemble agents from reusable blocks — models, rules, prompts, docs, and MCP tools — and share them with your team.

Key Features

• AI chat, autocomplete, and agent mode inside VS Code and JetBrains IDEs
• A CLI agent for the terminal and continuous integration pipelines
• Bring any model: OpenAI, Anthropic, Gemini, local models via Ollama, and more
• Continue Hub to build, version, and share agents and reusable blocks
• Rules, codebase context, and MCP tool integration for grounded answers
• Open-source core under the Apache-2.0 license

Continue is model-agnostic by design: point it at cloud APIs or at fully local models through Ollama or LM Studio for private, offline work. Its context system lets you attach files, docs, database schemas, and terminal output to a conversation with @-mentions, and team rules can be checked into the repository so every developer’s agent follows the same conventions.

💳 Pricing

Continue is freemium. The open-source extensions and CLI are free, and you pay only for your own model usage. Continue also offers paid Hub plans for teams that want hosted sharing, management, and collaboration features.

Who Is It For

Developers and engineering teams who want a capable AI coding agent without being locked into a single model vendor, and who value open-source tooling they can inspect, extend, and self-host.