Continue vs Cline

Where Cline wins

It is the one that is actually shipping

v4.1.11 was released on 2026-08-21, and Cline's release cadence is the fastest in this shortlist. Continue's last release was 2026-06-19 and its last commit on main was 2026-07-21. For a tool whose whole job is to broker between your editor and model APIs that change monthly, that difference is not cosmetic: a provider adds a model, a rate limit changes, an SDK deprecates something, and one of these two projects will respond.

The default is at-cost, and the paid tiers say so themselves

Cline's pricing page lists Open Source as free and Enterprise as custom. In between are two opt-in conveniences: pay-as-you-go Cline credits, and ClinePass at $9.99 per month standard rate, currently $4.99 for the first month, which the docs describe as 2 to 5 times the usage on popular open coding models compared with standard API rates. Those are vendor claims about their own gateway, so read them as marketing; the sentence worth quoting is Cline's own, that ClinePass is completely optional and you do not need it to use Cline. That is the same deal Continue offered, with the difference that nobody has turned the lights off.

The licence is real Apache-2.0, with no carve-out

Verified at the repository's LICENSE file: Apache-2.0 across the project, with no ee/ directory carrying a different licence. That matters because it is not the norm in this category. Tabby, for example, is Apache-2.0 except everything under ee/, which requires paid seats for production use, and Roo Code's successor Roomote went to the Fair Core License, Version 1.0, ALv2 Future License, which is source available rather than open source until each release converts to Apache-2.0 on its second anniversary.

It stays in your editor

Every other strong option here asks you to change how you work: goose is a desktop and terminal agent, opencode is a terminal agent, Zed is an entire editor. Cline is a VS Code extension first, so the migration is an install and a set of API keys rather than a new habit, and if you also want a terminal or scripted path it ships a GA CLI on npm, an SDK and a desktop build alongside the extension.

Where Continue still wins

Cline has not rebuilt Continue's config file, and that is the practical friction: your config.yaml does not port, and anything you had in Continue's Hub cannot be exported because that host has no DNS record and the export page returns 404. One concrete gap: Continue shipped inline tab completion as you type, and Cline does not. Cline is agent-only by design, and tab completion is still an open feature request in its tracker. If typing-time suggestions were the half of Continue you used daily, Cline replaces the chat and agent half only, and you will want a separate completion model for the rest.

And the honest structural point cuts Continue's way once, briefly. Continue's risk is now zero, because the thing that could go wrong already did: no company, no billing, no server to switch off, and an irrevocable licence on the code you have installed. Cline's risk is live. It is one company, one founder-CEO, $32M of institutional money that will eventually want a return, and no written governance saying who inherits the merge button. That is not an accusation, it is the same shape as every project on this page except goose, which is held by a foundation, and Zed, whose copyleft licence forces a fork to stay open. If your organisation now treats custody as a procurement question, Cline will not pass it. If your organisation wants software that gets patched next month, Continue will not pass that.

Migrating Continue to Cline

The path most teams take:

1. Inventory first. Pull your model definitions, rules and prompts out of config.yaml and out of your notes. Hub-sourced assistants and rules are unrecoverable; the 30-day export window closed on 2026-07-15 and continue.dev/export now 404s, so treat whatever is on your disk as the only copy.
2. Install and bring the keys, not the file. Add your provider keys to Cline, or point it at the local model you were already using. Skip credits and ClinePass on day one; the free path is the default and there is no reason to add a meter while you are still evaluating.
3. Run both for one cycle. Continue is not going to break while you decide, which is the one convenience of a shutdown that has already happened. Both tools bill through your own provider account, so the overlap costs extra tokens rather than a second subscription.
4. What does not move. Your Continue chat history, your Hub items, and whatever you scripted against api.continue.dev with the cn CLI or the Continue SDK, since that endpoint no longer answers. If that scripting is the core of your setup, look at opencode instead; if custody is the thing you cannot risk twice, read the goose comparison first.