Skip to main content

Claude

Anthropic Claude integration allows your product to connect with Anthropic's AI platform to use powerful large language models such as Claude 3, Claude 3.5 Sonnet, and Claude 3 Opus. This integration enables your product to perform text generation, summarization, coding assistance, conversational AI, and natural language understanding tasks.

Credentials Needed​

To integrate with Anthropic's API, you need an API Key associated with your Anthropic account.

Required credentials:

  • Anthropic API Key (e.g., sk-ant-xxxxxxxxxx)

API keys are account-specific and tied to your billing plan. Treat them as sensitive credentials and do not expose them in client-side code.

Permissions Needed / API Scopes​

Anthropic's API uses a single API key that provides access to all available capabilities under your account — there are no user-defined scopes.

FunctionalityEndpointDescription
Generate text or conversations/v1/messagesSend user messages and receive AI-generated responses
List available models/v1/modelsRetrieve the list of Claude models available to your account
Stream responses/v1/messages (with stream: true)Receive tokenized responses for real-time applications

Note: Anthropic does not currently offer per-feature scopes — all permissions are managed via the API key.

Creating Users / Access Tokens​

Step 1: Generate an API Key

  1. Log in to your Anthropic account: https://console.anthropic.com/
  2. Go to Settings → API Keys or directly visit: https://console.anthropic.com/settings/keys
  3. Click Create Key.
  4. Enter a recognizable name (e.g., ClaudeIntegrationKey).
  5. Copy the API key (sk-ant-xxxx) immediately — it will only be shown once.

Test Connectivity​

You can verify connectivity with the Claude API using curl or your programming environment:

curl https://api.anthropic.com/v1/messages \
  -H "x-api-key: <ANTHROPIC_API_KEY>" \
  -H "content-type: application/json" \
  -d '{
    "model": "claude-3-opus-20240229",
    "max_tokens": 100,
    "messages": [
      {"role": "user", "content": "Hello Claude! Please confirm connectivity."}
    ]
  }'

If you receive a valid JSON response with a model output, your credentials and permissions are configured correctly.

Save the Results in the Platform and Create Connection​

  1. In your platform's integration setup, securely store:
    • ANTHROPIC_API_KEY
  2. Create a connector labeled Anthropic Claude Integration.
  3. Test the connection by making a simple query to verify successful authentication.
  4. Optionally, allow users to specify their preferred Claude model (e.g., claude-3-opus, claude-3-sonnet, or claude-3-haiku).

Best Practices​

  • Store your API key securely using a secret manager or environment variable — never expose it in frontend applications.
  • Rotate keys regularly and immediately revoke unused or compromised keys.
  • Use rate limiting and error handling to manage API limits gracefully.
  • For streaming responses (chat-like experiences), use Anthropic's streaming API with WebSocket or SSE for optimal UX.
  • Log API usage and costs for monitoring and billing control.
  • Keep your integration flexible to support multiple Claude models (e.g., Opus, Sonnet, Haiku) depending on performance or cost needs.
  • For enterprise applications, use Anthropic's enterprise workspace for better access control and usage management.