--- title: Anthropic SDK product: vercel url: /docs/sandbox/ecosystem/anthropic-sdk canonical_url: "https://vercel.com/docs/sandbox/ecosystem/anthropic-sdk" last_updated: 2026-08-07 type: tutorial prerequisites: - /docs/sandbox/ecosystem - /docs/sandbox related: - /docs/sandbox - /docs/ai-gateway - /docs/cli - /docs/ai-gateway/authentication-and-byok/authentication - /docs/sandbox/sdk-reference summary: Use Anthropic tool use with AI Gateway to execute model-generated code in an isolated Vercel Sandbox. install_vercel_plugin: npx plugins add vercel/vercel-plugin --- # Anthropic SDK The [Anthropic SDK](https://github.com/anthropics/anthropic-sdk-typescript) supports tool use through the Messages API. This guide shows how to define a tool that executes code in [Vercel Sandbox](/docs/sandbox), so model-generated code runs in an isolated microVM instead of on your host. Model requests route through [AI Gateway](/docs/ai-gateway), and one set of Vercel credentials covers both the model and the 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. - [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 - [Using Vercel Sandbox to run Claude’s Agent SDK](https://vercel.com/kb/guide/using-vercel-sandbox-claude-agent-sdk) — Learn how to deploy Claude's Agent SDK in Vercel Sandbox for secure and isolated execution of AI-powered code generation - [Safely running AI generated code in your Next.js application](https://vercel.com/kb/guide/running-ai-generated-code-sandbox) — How to execute untrusted, AI‑generated code from a Next.js app using Vercel Sandbox, an isolated, ephemeral environment. - [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 - [Building an agent with OpenAI Agents SDK and Vercel Sandbox](https://vercel.com/kb/guide/building-an-agent-with-openai-agents-sdk-and-vercel-sandbox) — Learn how to build an agent with with OpenAI Agents SDK and Vercel Sandbox - [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. Full cross-link map for this page: [/docs/sandbox/ecosystem/anthropic-sdk.graph.md](/docs/sandbox/ecosystem/anthropic-sdk.graph.md) > **💡 Note:** You need a [Vercel account](https://vercel.com/signup), [Vercel > CLI](/docs/cli), and Node.js 22 or later. ## Getting started - ### Create a new project Create a new directory for your project and initialize it: ```bash filename="terminal" mkdir anthropic-sandbox cd anthropic-sandbox npm init -y ``` - ### Install dependencies Install the Anthropic SDK and the Sandbox SDK: **pnpm** ```bash pnpm i @anthropic-ai/sdk @vercel/sandbox ``` **yarn** ```bash yarn add @anthropic-ai/sdk @vercel/sandbox ``` **npm** ```bash npm i @anthropic-ai/sdk @vercel/sandbox ``` **bun** ```bash bun add @anthropic-ai/sdk @vercel/sandbox ``` - ### Configure authentication Link the project to Vercel and pull your development environment. This creates a `.env.local` file with a `VERCEL_OIDC_TOKEN`, which authenticates both the Sandbox SDK and AI Gateway: ```bash filename="terminal" vercel link vercel env pull ``` > **💡 Note:** You can also use an [AI Gateway API > key](/docs/ai-gateway/authentication-and-byok/authentication) by setting > `AI_GATEWAY_API_KEY`. > For other Sandbox authentication methods, see > [Authentication](/docs/sandbox/sdk-reference#authentication) in the SDK reference. - ### Create your application Create a new file called `index.ts` with the following code: ```typescript filename="index.ts" import Anthropic from '@anthropic-ai/sdk'; import { Sandbox } from '@vercel/sandbox'; const gatewayApiKey = process.env.AI_GATEWAY_API_KEY; const anthropic = new Anthropic({ ...(gatewayApiKey ? { apiKey: gatewayApiKey } : { authToken: process.env.VERCEL_OIDC_TOKEN }), baseURL: 'https://ai-gateway.vercel.sh', }); const sandbox = await Sandbox.create({ timeout: 300_000 }); try { const tools: Anthropic.Tool[] = [ { name: 'run_code', description: 'Run Node.js code in an isolated Vercel Sandbox and return its output.', input_schema: { type: 'object', properties: { code: { type: 'string', description: 'Node.js code to execute' }, }, required: ['code'], }, }, ]; const messages: Anthropic.MessageParam[] = [ { role: 'user', content: "Compute the SHA-256 hex digest of the exact string 'vercel sandbox' with Node.js, then report the digest.", }, ]; let response = await anthropic.messages.create({ model: 'anthropic/claude-opus-5', max_tokens: 16000, tools, messages, }); while (response.stop_reason === 'tool_use') { messages.push({ role: 'assistant', content: response.content }); const toolResults: Anthropic.ToolResultBlockParam[] = []; for (const block of response.content) { if (block.type !== 'tool_use') continue; const { code } = block.input as { code: string }; const result = await sandbox.runCommand('node', ['-e', code]); const output = result.exitCode === 0 ? await result.stdout() : `Command failed: ${await result.stderr()}`; toolResults.push({ type: 'tool_result', tool_use_id: block.id, content: output, }); } messages.push({ role: 'user', content: toolResults }); response = await anthropic.messages.create({ model: 'anthropic/claude-opus-5', max_tokens: 16000, tools, messages, }); } for (const block of response.content) { if (block.type === 'text') console.log(block.text); } } finally { await sandbox.stop(); } ``` The `timeout` value bounds each sandbox session to five minutes. Model calls run outside the VM, and a later `sandbox.runCommand()` can resume a new session after the previous one times out. The session timeout doesn't bound the overall agent loop. Use an application deadline and cancellation strategy to bound the full loop, including model requests and tool submissions. A command's `timeoutMs` (or Python `kill_after`) bounds that command separately. The `finally` block stops the current session when the loop finishes or throws; coordinate cancellation with in-flight tool calls before cleanup so they don't resume another session. The following code: - Points the Anthropic SDK at AI Gateway, authenticated with your Vercel credentials - Creates a sandbox that is reused across tool calls - Declares a `run_code` tool the model can call - Executes each tool call with [`sandbox.runCommand()`](/docs/sandbox/sdk-reference#sandboxruncommand) and returns the output as a `tool_result` block - Loops while `stop_reason` is `tool_use`, then prints the final answer - ### Run the application Run your application with the environment variables from `.env.local`: ```bash filename="terminal" npx tsx --env-file=.env.local index.ts ``` Claude writes Node.js code, the sandbox executes it, and the final response reports the digest. --- [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.