Continue.dev AI coding — Open-source, model-agnostic AI code assistant for VS Code, JetBrains and CLI - frozen after its acquisition by Cursor.
What it is
Continue is a free, Apache-2.0 AI coding assistant offering chat, autocomplete and inline edits in VS Code, JetBrains IDEs and a CLI. Cursor acquired the company and development has ended.
Editor's review
Long-form introduction by the BetterPicker editors · checked against the official site · Oct 23, 2026
Continue was the model-agnostic answer to lock-in: an open-source AI code assistant with a chat sidebar, autocomplete and inline edits inside VS Code, JetBrains IDEs and a CLI, working with whichever provider you plugged in - Anthropic, Mistral, OpenAI, or fully local models through Ollama - so output quality tracked your model choice rather than a vendor's. That chapter has closed. Cursor acquired Continue, the official site is now a farewell note, and the GitHub repository is read-only after a final 2.0.0 release of the extension, plugin and CLI. The judgment cuts both ways. As free Apache-2.0 software whose last release removed anonymous telemetry and authentication requirements, it remains a genuinely useful, inspectable tool that costs nothing beyond your own model keys. As a maintained product with fixes and day-one support for new models, it is finished. Adopt it as a foundation to fork and own, not a companion that will grow with you.
What it does well
Total model freedom: connect Anthropic, Mistral, OpenAI or local models through Ollama and LM Studio, then switch per task; reviewers consistently found it worked with more providers than competing extensions, which is the reason many installed it.
A complete assistant surface: chat with file and folder references, autocomplete that closes tags and continues code, inline editing directly in the editor, and slash commands for writing documentation, comments, shell commands or unit tests - with custom prompts as a first-class feature.
Free and inspectable: Apache-2.0 licensed and free to run indefinitely with your own keys, and the final 2.0.0 release stripped anonymous telemetry and sign-in requirements, leaving a clean codebase explicitly donated to the community as a foundation.
Who it's for
Privacy-conscious developers, local-model hobbyists already running Ollama, and teams that want to fork an assistant and own it outright rather than rent one per seat. It also suits anyone who already pays for model APIs elsewhere and resents paying a second subscription on top. If you expect active maintenance, immediate support for newly released models, or agentic features like terminal control, choose one of the actively developed alternatives instead.
Where it falls short
Development has ended: the Cursor acquisition froze the project; the repository is read-only, the JetBrains plugin is already deprioritized in favor of the CLI, and future IDE updates or model APIs will go unpatched.
Context is manual labor: automatic codebase awareness is absent, so reviewers found themselves attaching files and folders by hand for every request and noted that rival agents discovered relevant code on their own.
Setup rides on configuration: walkthroughs show friction editing the JSON config, juggling separate API keys per provider, and working around occasional shortcut quirks - acceptable for tinkerers, punishing for anyone who wants an instant install.
Specs at a glance
Facts from the official site · not editorial opinion
| Latest version | 2.0.0 (final release of the VS Code extension, JetBrains plugin and CLI) |
|---|---|
| Price | Free and open source; you supply your own model access or run models locally |
| Platforms | VS Code, JetBrains IDEs, CLI (macOS, Linux, Windows) |
| Open source | Yes - Apache-2.0 |
| Core capability | Bring-your-own-model AI assistant with chat, autocomplete and inline edits |
| Integrations | Anthropic, Mistral, OpenAI, Ollama, LM Studio and other providers; file, folder and codebase references |
| Best for | Developers who want a model-agnostic, forkable assistant and accept that maintenance has ended |
Frequently asked questions
▸Is Continue free to use?
Yes. It is Apache-2.0 open source with no subscription. You connect your own API keys from providers like Anthropic, Mistral or OpenAI, or run models locally through Ollama, so your only cost is model usage.
▸Can Continue work with local models?
Yes. Local models through Ollama and LM Studio are first-class options, which made Continue a common pick for developers who want AI assistance without sending code to external services.
▸What happened to Continue?
Cursor acquired the company. The official site is now a farewell announcement, the GitHub repository is read-only, and a final 2.0.0 release shipped with telemetry removed and authentication pulled out before development stopped.
▸How does Continue compare with Cline or GitHub Copilot?
Reviewers credit Continue with broader provider choice and stronger custom prompts, while Cline offers terminal access and automatic codebase awareness that Continue lacks, and Copilot provides a maintained subscription with vendor-managed models rather than bring-your-own flexibility.
Video reviews
3 review videos aggregated · praise and criticism included alike · click through to the original video
Channels that covered it
Channels are aggregated as sources only — we don’t rate creators
Related tools
Where to go next
External links open in a new tab; external content is independent of this site.
Link down? Every object page is re-checked monthly.