Can you use your Claude plan in OpenCode or other tools?

No, not inside OpenCode. Anthropic's Claude Code legal page says third-party developers may not offer Claude.ai login in their apps or route requests through Free, Pro or Max plan credentials, and OpenCode stopped bundling the plugins that did it in version 1.3.0. You can still use Claude in OpenCode with an API key, Amazon Bedrock or Google Vertex AI, or use your Claude subscription in Claude Code itself.

We read Anthropic's page, OpenCode's docs and Cline's docs on 24 October 2026 and quote them below. This is our reading of them, not legal advice, and Anthropic is the one who enforces its terms.

What Anthropic's terms say about Claude subscriptions in other tools#

The rules sit in the "Authentication and credential use" section of Claude Code's legal and compliance page. It starts with who the subscription sign-in is for:

OAuth authentication is intended exclusively for purchasers of Claude Free, Pro, Max, Team, and Enterprise subscription plans and is designed to support ordinary use of Claude Code and other native Anthropic applications.

Then it turns to everyone who builds a product on Claude, with the Agent SDK named among them, and tells them to use an API key from Claude Console or a supported cloud provider. The next two sentences answer the OpenCode question:

Anthropic does not permit third-party developers to offer Claude.ai login into their own applications, or to route requests through Free, Pro, or Max plan credentials on behalf of their users. Moreover, developers may not collect, store, or intermediate Claude.ai credentials or session tokens — sign-in to a Claude account must complete through Anthropic's own flow.

Two things stay allowed. Customers may provision their own API keys and cloud-provider credentials for their own authorised users, as long as the usage is billed to the key owner and isn't resold. And the rules don't prevent "an end user from signing in to the unmodified Claude Code binary with their own Claude subscription", including where a platform hosts Claude Code for you. Then comes the warning: Anthropic "reserves the right to take measures to enforce these restrictions and may do so without prior notice".

Free, Pro and Max fall under the Consumer Terms; Team, Enterprise and the API fall under the Commercial Terms, per the same page. What the Consumer Terms say about sharing a login and holding several accounts is quoted in our guide to the rules for multiple Claude accounts.

Why OpenCode no longer works with Claude Pro and Max#

OpenCode used to ship plugins that signed in with a Claude subscription. Its providers page now says, under Anthropic:

There are plugins that allow you to use your Claude Pro/Max models with OpenCode. Anthropic explicitly prohibits this. Previous versions of OpenCode came bundled with these plugins but that is no longer the case as of 1.3.0

OpenCode 1.3.0 came out on 22 March 2026, and the first item in its release notes is "Removed anthropic oauth plugin". Third-party plugins still exist, as OpenCode's page says, and installing one is the setup Anthropic's sentence about routing requests through Pro or Max credentials rules out. The same providers page lists ChatGPT Plus, GitHub Copilot and GitLab Duo as subscriptions that work in OpenCode with no setup; Claude isn't on that list.

Cline, Roo Code and tools that run Claude Code for you#

A second pattern doesn't hand your Claude sign-in to the other app. Cline's docs describe a Claude Code provider: you install the claude CLI and sign in with Anthropic's own flow, then pick Claude Code as Cline's provider and give it the path to claude. Cline says this "uses your Claude subscription limits instead of API token billing". Roo Code had a Claude Code provider too, and its 3.42.0 release on 22 January 2026 removed it.

Anthropic's page doesn't name this setup. It allows signing in to the unmodified binary with your own plan, and it tells developers building products to use API keys and not to route requests through Free, Pro or Max credentials for their users. Which sentence a given tool falls under is Anthropic's call, and we won't make it on Anthropic's behalf.

Using a claude setup-token token somewhere else#

claude setup-token prints a one-year token for your subscription. Anthropic's authentication docs describe it "for CI pipelines and scripts where browser login isn't available", set as CLAUDE_CODE_OAUTH_TOKEN for Claude Code, and say it "can only make model requests". Pasting it into an app that isn't Claude Code means that app sends requests on your plan's credentials. Our claude setup-token guide covers what the token is for.

What is allowed, at a glance#

How to use Claude in OpenCode the allowed way#

1. Create an API key on the API keys page of Claude Console.
2. In OpenCode, run /connect and select Anthropic.
3. Paste the key when OpenCode asks for it.
4. Run /models and pick a Claude model.

A key bills every token and has no 5-hour or weekly window. Our API key vs subscription guide prices the two side by side. If your company already buys Claude through AWS or Google Cloud, OpenCode's providers page also covers Amazon Bedrock and Google Vertex AI; the same route for Claude Code itself is in Claude Code with Bedrock or Vertex AI.

If what you want is your Max plan for Claude and OpenCode for other models, the plain answer is to run both: Claude Code on the plan, OpenCode on its own providers. How the two fit next to each other is in our OpenCode comparison.

How Shuriken handles a Claude sign-in#

Shuriken runs the real Claude Code. A Claude Code chat starts the claude program installed on that computer, which Shuriken installs from Anthropic's own downloads when it's missing, and drives it through Anthropic's Agent SDK library. We say that plainly because the page quoted above names the Agent SDK in its sentence about API keys.

When you add a Claude account as a source, Shuriken shows Anthropic's sign-in link. You sign in on Anthropic's own page, and if it gives you a code, you paste the code into Shuriken. The claude program doesn't finish that sign-in: a small script of Shuriken's does, talking to Anthropic's sign-in servers itself under Claude Code's client ID, and it leaves the result in a Claude Code folder of that source's own. Shuriken moves the sign-in from there into its own encrypted store on that computer, and copies it to your other devices through your vault, end-to-end encrypted, so our cloud holds only ciphertext and is never sent a Claude sign-in it can read. From then on Shuriken refreshes the sign-in itself. Each Claude Code chat starts with the current access token and sends its requests through a small proxy Shuriken runs on the same computer, which sends each one on with the token of the source the chat is on at that moment.

The same page says developers may not "collect, store, or intermediate Claude.ai credentials or session tokens", and that "sign-in to a Claude account must complete through Anthropic's own flow". Whether a sign-in finished by our script, kept, refreshed and synced by Shuriken, with Claude Code's requests passed on through its proxy, fits those sentences is Anthropic's question to answer, not ours, and we don't claim its approval for any of this. If you would rather keep your plan out of that question, add a Claude Console API key as a source instead, or use Claude Code on its own. Whatever you add, add only accounts you own and are entitled to use.

Shuriken also runs OpenCode as an agent of its own, with OpenCode's own sign-in, so an OpenCode chat there uses whatever providers you set up in OpenCode. If that is the setup you want, download Shuriken.