Continue

What it does

Continue is an open-source AI coding assistant that integrates into VS Code and JetBrains IDEs - providing inline code completion, chat-based code generation, and codebase context-aware assistance. Its defining feature is model flexibility: Continue works with any LLM provider including OpenAI, Anthropic Claude, Google Gemini, and locally-running models via Ollama - giving developers full control over which AI powers their assistance without being locked into a single vendor. It is configured via a simple JSON file, supports custom prompts, and can connect to documentation, codebase indexing, and custom context sources. Continue is popular among developers who want AI coding assistance with privacy controls, local model support, or cost control over API usage.

Strengths

• Mid-market engineering organizations use Continue for team-wide AI coding assistance with centralized model configuration - policy-driven model selection and shared prompt libraries across the engineering team.
• Development teams use Continue for AI-assisted coding with enterprise model flexibility - connecting to internally approved LLM providers or local models that satisfy data governance requirements.
• Small engineering teams use Continue as a cost-effective AI coding assistant - open-source with no per-seat license fees, each developer controlling their own LLM provider and API costs.
• Individual developers use Continue for AI coding assistance with full model choice - local models for privacy, Claude for quality, or switching between providers based on task needs.

Watch-outs

• Requires developer configuration: Continue is more technical to set up than commercial AI coding tools like Cursor or GitHub Copilot — developers must configure LLM providers, API keys, and context settings, which is a barrier for non-technical users.
• No managed cloud infrastructure: Continue is a self-configured extension — teams wanting managed enterprise features (centralized billing, admin controls, usage analytics) need to build those processes themselves or choose a commercial alternative.
• Quality depends on chosen LLM: Continue is a framework — the quality of code assistance depends entirely on the underlying model chosen. Selecting and paying for a capable LLM (Claude, GPT-4) is necessary to match commercial AI coding tool quality.

Pricing

Continue itself is free and open-source. Users pay LLM API providers directly. {{claude}} API approximately $3 to $15 per million tokens. {{chatgpt}}-4 at comparable rates. Local models via Ollama are free to run on local hardware.