Open-source, cross-platform AI chat client for Web, desktop and mobile that connects to OpenAI, Claude, Gemini, DeepSeek and self-hosted models.
anthropics/anthropic-sdk-typescript
Official TypeScript SDK for calling Anthropic's Claude API from server-side JavaScript and TypeScript applications.
- Stars
- 2,147
- Forks
- 427
- Open issues
- 98
- Last commit
- 15 hours ago
What is it?
This is the official TypeScript and JavaScript client library for Anthropic's Claude API. It wraps the HTTP API in a typed client so applications can send messages to Claude models, install it with npm, and receive responses without hand-writing requests.
The SDK covers streaming responses, structured outputs, tool helpers and a tool runner, with separate guides for browser and computer toolsets. It also reports each API call as an OpenTelemetry span when a tracer provider is registered, with options to change content capture and trace propagation.
It runs on Node.js 20 and later, Deno, Bun, Cloudflare Workers, Vercel Edge Runtime, Nitro and Jest's node environment. Browser use is disabled by default to avoid exposing API keys and must be enabled explicitly. React Native is not supported.
Key features
- 01Typed client for the Claude Messages API with an npm install.
- 02Streaming responses, structured outputs and tool helpers.
- 03Tool runner for executing tool calls within a conversation.
- 04OpenTelemetry spans for each API call, with configurable content capture.
- 05Runs on Node.js, Deno, Bun, Cloudflare Workers and Vercel Edge Runtime.
Use cases
- Adding Claude-powered chat or text generation to a Node.js backend.
- Building serverless or edge functions that call Claude.
- Implementing tool-using agents with the built-in tool runner.
- Tracing LLM calls in an application that already uses OpenTelemetry.
Quick start
- 1
Install the SDK from npm
npm install @anthropic-ai/sdk - 2
Set your ANTHROPIC_API_KEY environment variable
- 3
Create a client and call client.messages.create in your app
Things to know
- Calling the Claude API requires an Anthropic API key.
- Browser use is disabled by default to protect API credentials and must be enabled explicitly.
- React Native is not supported at this time.
Topics
FAQ
- What is anthropic-sdk-typescript?
- Official TypeScript SDK for calling Anthropic's Claude API from server-side JavaScript and TypeScript applications. This is the official TypeScript and JavaScript client library for Anthropic's Claude API. It wraps the HTTP API in a typed client so applications can send messages to Claude models, install it with npm, and receive responses without hand-writing requests.
- Is anthropic-sdk-typescript free to use? What license does it have?
- anthropic-sdk-typescript is open source under the MIT license. Check the license terms before using it in commercial products.
- Is anthropic-sdk-typescript still maintained?
- Actively maintained. The last commit to anthropic-sdk-typescript was on Oct 20, 2026.
- What can I use anthropic-sdk-typescript for?
- Adding Claude-powered chat or text generation to a Node.js backend. Building serverless or edge functions that call Claude. Implementing tool-using agents with the built-in tool runner. Tracing LLM calls in an application that already uses OpenTelemetry.
- How do I install anthropic-sdk-typescript?
- Install the SDK from npm: npm install @anthropic-ai/sdk Set your ANTHROPIC_API_KEY environment variable Create a client and call client.messages.create in your app
- Where can I get anthropic-sdk-typescript?
- The source code and installation instructions are on GitHub: https://github.com/anthropics/anthropic-sdk-typescript
Summary generated by AI from the project's README. GitHub stats refresh daily.
Related projects
Open source autonomous coding agent that runs as an IDE extension, terminal CLI, desktop app, or SDK.
A TypeScript coding agent CLI where a single 'ulw' or 'mass ulw' keyword in your prompt makes it plan, delegate to multiple models, and verify the work.
The official Python library for the OpenAI REST API, with sync and async clients and typed request and response models.
TypeScript and Python SDK giving AI agents 1000+ pre-authenticated app toolkits, per-user sessions, and a sandbox to take actions.
The official TypeScript SDK for building Model Context Protocol servers and clients that give LLM applications standardized context.