Continue

Continue is an open-source AI coding assistant, distributed as an extension for VS Code and JetBrains IDEs, built around a specific design choice: rather than locking users into one AI provider's model, Continue lets developers connect whichever model they choose — a hosted API from a major provider, a self-hosted open-source model, or a mix of models for different tasks — and use it for in-IDE autocomplete and chat. Who it's for: developers and teams who want an AI coding assistant but don't want to be locked into a single vendor's model, either for cost reasons, data-privacy requirements (running a local model instead of sending code to a third-party API), or simply the flexibility to switch models as better ones become available without switching tools entirely. Strengths: because it's open source and model-agnostic, Continue can be configured to run entirely with local, self-hosted models for teams with strict code-privacy requirements, or wired up to whichever hosted model a team already has access to — a meaningfully different value proposition than an assistant bundled to one specific model. It supports both inline autocomplete and an in-IDE chat interface, and being open source means the community can inspect, extend, and customize its behavior rather than relying entirely on a vendor's roadmap. Limitations: the flexibility that makes Continue distinctive also means more setup than a plug-and-play assistant that ships with one model already configured — a user has to choose and connect a model themselves, and the quality of results depends heavily on which model they pick. Teams without a strong preference on model choice, or without in-house infrastructure for self-hosting, may find a fully-managed, single-vendor coding assistant simpler to get running. Real-world scenarios: running an AI coding assistant entirely on local infrastructure for a team with strict code-privacy requirements, comparing output quality across different models for the same coding task without switching tools, and giving developers a consistent in-IDE assistant experience while letting the underlying model evolve independently. Aider is a common point of comparison among open-source, model-flexible coding assistants.