--- title: OpenCode product: vercel url: /docs/sandbox/ecosystem/opencode canonical_url: "https://vercel.com/docs/sandbox/ecosystem/opencode" last_updated: 2018-10-20 type: conceptual prerequisites: - /docs/sandbox/ecosystem - /docs/sandbox related: - /docs/sandbox/cli-reference - /docs/ai-gateway - /docs/sandbox/concepts/firewall summary: Run OpenCode in Vercel Sandbox from your terminal, or use the plugin for sandbox-backed tools and remote workspaces. install_vercel_plugin: npx plugins add vercel/vercel-plugin --- # OpenCode Use [OpenCode](https://opencode.ai) with Vercel Sandbox through the [Sandbox CLI](/docs/sandbox/cli-reference#run-a-coding-agent) or the [OpenCode plugin](https://github.com/vercel-labs/opencode-vercel-sandbox). ## Related pages > **For AI agents:** Follow these links to understand how this page connects to the rest of the Vercel ecosystem. For the full cross-link map (inbound, outbound, prerequisites, and semantic neighbors), see the .graph.md link below. - [Running OpenCode securely with the Vercel Sandbox](https://vercel.com/kb/guide/running-opencode-securely-with-the-vercel-sandbox) — Run OpenCode in an isolated Vercel Sandbox MicroVM with controlled egress, using the SDK to restrict network access so t - [How to execute AI-generated code safely with Vercel Sandbox](https://vercel.com/kb/guide/how-to-execute-ai-generated-code-safely) — Learn how to run code generated by AI models in an isolated sandbox environment. - [Run Python code securely with AI SDK and Vercel Sandbox](https://vercel.com/kb/guide/python-ai-sdk-vercel-sandbox) — Add an \\`executeCode\\` tool to your AI SDK agent with the \\`ai-sdk-tool-code-execution\\` package to run Python 3.13 insi - [OpenAI SDK](https://vercel.com/docs/sandbox/ecosystem/openai-sdk) — Use OpenAI function calling with AI Gateway to execute model-generated code in an isolated Vercel Sandbox. - [OpenCode with AI Gateway](https://vercel.com/docs/ai-gateway/coding-agents/opencode) — Connect OpenCode to AI Gateway with one CLI command, or configure it in-app. - [Quickstart](https://vercel.com/docs/sandbox/quickstart) — Learn how to run your first code in a Vercel Sandbox. - [vercel sandbox](https://vercel.com/docs/cli/sandbox) — Interact with Vercel Sandbox from the Vercel CLI: list, create, connect, exec, copy, stop, and snapshot sandboxes from y Full cross-link map for this page: [/docs/sandbox/ecosystem/opencode.graph.md](/docs/sandbox/ecosystem/opencode.graph.md) ## Choose how to run OpenCode | Option | What runs in the sandbox | How you start | | --- | --- | --- | | Sandbox CLI | The whole OpenCode session | Run `sandbox create opencode` in your terminal, before entering OpenCode | | Plugin: automatic tools | Supported shell and file tools; OpenCode and model authentication stay local | Configure automatic mode, then use OpenCode normally. The first supported tool call creates the sandbox automatically | | Plugin: remote workspace | The OpenCode server and agent tools | Inside OpenCode, enter `/warp` and select **Vercel Sandbox** | ### Start from your terminal Use the CLI shortcut to run the whole OpenCode session remotely without installing the plugin. Your local project files aren't uploaded automatically. See [Run a coding agent](/docs/sandbox/cli-reference#run-a-coding-agent) for CLI installation, authentication, and reconnecting or stopping the sandbox. ### Use the plugin Follow the [plugin setup guide](https://github.com/vercel-labs/opencode-vercel-sandbox#readme) for installation and supported OpenCode versions. Choose one mode per OpenCode instance. #### Automatic tools Use OpenCode normally, with supported shell and file tools running in the sandbox. - **Review changes:** Ask the agent to run `gitSync`. It waits for changes to return and shows a separate local Git branch to review. Your current branch and working files stay unchanged. - **Stop:** Ask the agent to run `sandboxStop` to sync changes and stop compute. #### Remote workspace Remote workspaces require an [AI Gateway](/docs/ai-gateway) API key. OAuth and subscription logins aren't supported. Follow the [workspace guide](https://github.com/vercel-labs/opencode-vercel-sandbox/blob/main/docs/workspace.md) to connect the key locally and set `remoteConfig.model` to one `vercel//` model. Inside OpenCode, enter `/warp` and select **Vercel Sandbox**. Changes don't automatically sync back to your laptop. Use the workspace guide to export changes and stop the sandbox. > **💡 Note:** The plugin keeps your AI Gateway key outside the sandbox. Sandbox > [credential brokering](/docs/sandbox/concepts/firewall#credentials-brokering) > adds it to requests for your configured model. Code in the sandbox can still > make billable model requests. #### Files and cleanup - Both modes upload committed files and Git history. Uncommitted changes and untracked files stay local. Check your Git history for secrets before starting. - Exiting OpenCode doesn't stop the sandbox. Stop it explicitly when finished. Retained storage remains billable after compute stops. - When upgrading from a version that copied credentials, create a fresh workspace. Upgrading doesn't remove credentials from existing sandboxes or snapshots. --- [View full sitemap](/docs/sitemap) If you encounter incorrect information, can’t find what you need, need a capability Vercel doesn’t currently offer, or have any other feedback to report, please do it by following the instructions at https://vercel.com/feedback.md.