Install the Anthropic SDK

The Anthropic SDK is the official client library for the Claude API. It's published for Python (anthropic) and Node (@anthropic-ai/sdk) — same API surface in both. This post installs both and verifies each with a one-line call.

You'll need an API key from console.anthropic.com. Free credits cover the verify step many times over.

Python

Install via uv inside a project:

# add the anthropic sdk to your project

uv add anthropic

For a quick one-off without a project, use uv tool install or uv run:

# run a script with the sdk available, no project setup

uv run --with anthropic python -c "from anthropic import Anthropic; print(Anthropic().messages.create(model='claude-haiku-4-5-20251001', max_tokens=20, messages=[{'role':'user','content':'hi'}]).content[0].text)"

If you have an older project still using pip, pip install anthropic works fine.

Node

Install via npm inside a project:

# add the anthropic sdk to your project

npm install @anthropic-ai/sdk

Works identically with pnpm and yarn (pnpm add @anthropic-ai/sdk, yarn add @anthropic-ai/sdk).

Configure the API key

The SDK reads ANTHROPIC_API_KEY from the environment. Don't hardcode it. For development, put it in a .env file (and add .env to .gitignore):

# create .env with your api key

echo 'ANTHROPIC_API_KEY=sk-ant-...' >> .env

In Python, load it with python-dotenv or just export ANTHROPIC_API_KEY=... in your shell. In Node, load it with dotenv or rely on the shell env.

For production, use whatever secret manager your platform provides (AWS Secrets Manager, GCP Secret Manager, Vercel env vars). Never commit the key.

Verify (Python)

# verify python sdk with a one-line call

from anthropic import Anthropic; print(Anthropic().messages.create(model="claude-haiku-4-5-20251001", max_tokens=64, messages=[{"role":"user","content":"hi"}]).content[0].text)

Save to verify.py, then uv run python verify.py. You should see Claude's reply printed.

Verify (Node)

// verify node sdk with a one-line call

import Anthropic from "@anthropic-ai/sdk"; const r = await new Anthropic().messages.create({ model: "claude-haiku-4-5-20251001", max_tokens: 64, messages: [{ role: "user", content: "hi" }] }); console.log(r.content[0].text);

Save to verify.mjs, then node verify.mjs.

Common gotchas

• Wrong env var name: it's ANTHROPIC_API_KEY, not CLAUDE_API_KEY or ANTHROPIC_KEY. The SDK only reads the official name.
• Model IDs change: pin the full ID with the date suffix (claude-haiku-4-5-20251001) when you want stability. The shorter alias (claude-haiku-4-5) follows the latest snapshot and can change behavior under you.
• Rate limits hit fast on free credits: a tight loop of test calls will hit the per-minute limit. Add time.sleep(1) or use the usage field to back off.
• Streaming differs from one-shot: .messages.create is one-shot. For token-by-token streaming, use .messages.stream (Python) or .messages.stream (Node) — different return type, you iterate over deltas.
• Prompt caching: enable it explicitly with cache_control blocks. The SDK doesn't cache automatically; you tag what's cacheable. Worth it for any prompt over 1k input tokens that you reuse.

With verify.py or verify.mjs printing a reply, you're ready to build against Claude. Most production code differs from the verify script only in the message list and error handling.