MCP Servers

How to Get Your Anthropic API Key (Claude API)

Step-by-step guide to getting an Anthropic API key for the Claude API. Covers account setup, key generation, billing, security best practices, and first API call.

What Is an Anthropic API Key?

An Anthropic API key is a secret token that authenticates your application with the Claude API. Every request you send to Claude's models — whether for text generation, analysis, tool use, or embedding — must include this key in the request headers. The key ties your usage to your account for billing and rate-limiting purposes.

Anthropic offers API access to all Claude models including Claude Opus 4, Claude Sonnet 4, and Claude Haiku. The API is separate from the consumer Claude.ai chat interface. While Claude.ai uses a subscription model (free, Pro, or Team), the API charges per token — you pay only for what you use, making it the preferred option for developers building applications, running batch analysis, or integrating Claude into automated workflows.

Step 1: Create an Anthropic Account

Go to console.anthropic.com and sign up for an account. You can use email, Google, or GitHub authentication. This is the Anthropic Console — the developer dashboard separate from claude.ai. If you already have a claude.ai account, you still need to register for Console access separately.

After verifying your email, you will land on the Console dashboard. The dashboard shows your API usage, billing status, and key management interface. New accounts typically receive a small amount of free credits for testing.

Step 2: Add Billing Information

Navigate to the Billing section in the Console sidebar. Add a payment method (credit card or prepaid credits). Anthropic uses a pay-as-you-go model — you are charged based on the number of input and output tokens your API calls consume. You can set spending limits to prevent unexpected charges.

For testing and development, the costs are very low. A typical development session generating a few thousand responses might cost a few dollars. Production applications with heavy usage will want to monitor costs through the Console's usage dashboard and set appropriate alerts.

Step 3: Generate Your API Key

In the Console, go to API Keys (in the sidebar under your organization). Click "Create Key" and give it a descriptive name like "development" or "production-backend". The key will be displayed once — copy it immediately and store it securely. You cannot retrieve the full key later; you can only see the last few characters for identification.

# Your API key looks like this: sk-ant-api03-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx # Store it as an environment variable, not in code: export ANTHROPIC_API_KEY="sk-ant-api03-your-key-here"

You can create multiple keys for different environments (development, staging, production) or different applications. Each key can be individually revoked without affecting others, which is important for key rotation and security incident response.

Step 4: Make Your First API Call

Test your key with a simple curl command or Python script. Here are both approaches:

# Using curl curl https://api.anthropic.com/v1/messages \ -H "content-type: application/json" \ -H "x-api-key: $ANTHROPIC_API_KEY" \ -H "anthropic-version: 2023-06-01" \ -d '{ "model": "claude-sonnet-4-20250514", "max_tokens": 256, "messages": [ {"role": "user", "content": "Hello, Claude!"} ] }'
# Using Python pip install anthropic import anthropic client = anthropic.Anthropic() # reads ANTHROPIC_API_KEY env var message = client.messages.create( model="claude-sonnet-4-20250514", max_tokens=256, messages=[ {"role": "user", "content": "Hello, Claude!"} ] ) print(message.content[0].text)

If you get a successful response, your key is working. Common errors include 401 (invalid key), 403 (billing not set up), and 429 (rate limit exceeded). The Anthropic Python SDK automatically reads the ANTHROPIC_API_KEY environment variable, so you do not need to pass it explicitly in code.

API Key Security Best Practices

Never hardcode your API key in source files. Use environment variables, secrets managers (AWS Secrets Manager, HashiCorp Vault, 1Password), or your platform's built-in secrets handling (Vercel Environment Variables, Railway Variables, Heroku Config Vars). If a key is accidentally exposed in a Git commit, revoke it immediately through the Console and generate a new one — treat any exposed key as compromised.

For team environments, use organization-level keys with appropriate permissions rather than sharing personal keys. Anthropic's Console supports workspaces where team members can manage their own keys under a shared billing account. Set up spending alerts and rate limits per key to prevent abuse or accidental runaway costs.

Rotate API keys periodically, especially for production applications. Create the new key, update your deployment to use it, verify it works, then revoke the old key. This can be automated as part of your deployment pipeline.

Claude API Key vs Anthropic API Key

These terms refer to the same thing. "Anthropic API key" is the official name since Anthropic is the company that provides the API. "Claude API key" is the common informal name since Claude is the model you access through the API. Both terms lead to the same Console at console.anthropic.com and the same key format starting with sk-ant-.

Note that API keys for third-party services that integrate Claude (like Amazon Bedrock or Google Cloud Vertex AI) are different. Those platforms use their own authentication mechanisms. An Anthropic API key only works with Anthropic's direct API at api.anthropic.com.

Using Your API Key with MCP Servers

If you are building MCP servers that call the Claude API internally (for example, an MCP server that uses Claude to analyze or summarize data before returning results), pass the API key as an environment variable in the server configuration. This keeps the key secure while making it available to the server process. See our Python MCP server tutorial for implementation patterns.

Last updated: 2026 • Browse all courses