Quickstart

Prerequisites

Before you begin, make sure you have:

An Anthropic account at console.anthropic.com
An API key from the API Keys page
Postman or a terminal with curl available

Never hardcode your API key in any file you commit to version control. Store it as an environment variable and reference it from there.

Your First API Call

You will send a message to Claude and receive a response. This is the foundation of every interaction with the Claude API.

Using Postman

Set up your request with the following: Method: POST URL: https://api.anthropic.com/v1/messages Headers:

Key	Value
`x-api-key`	`YOUR_API_KEY`
`anthropic-version`	`2023-06-01`
`content-type`	`application/json`

Body:

{
  "model": "claude-sonnet-4-6",
  "max_tokens": 1024,
  "messages": [
    {
      "role": "user",
      "content": "Hello, Claude. Respond in one sentence."
    }
  ]
}

Click Send.

Using curl

curl https://api.anthropic.com/v1/messages \
  -H "x-api-key: $ANTHROPIC_API_KEY" \
  -H "anthropic-version: 2023-06-01" \
  -H "content-type: application/json" \
  -d '{
    "model": "claude-sonnet-4-6",
    "max_tokens": 1024,
    "messages": [{"role": "user", "content": "Hello, Claude. Respond in one sentence."}]
  }'

Understanding the Response

A successful request returns a 200 OK with a response body like this:

{
  "id": "msg_01XFDUDYJgAACzvnptvVoYEL",
  "type": "message",
  "role": "assistant",
  "content": [
    {
      "type": "text",
      "text": "Hello! How can I help you today?"
    }
  ],
  "model": "claude-sonnet-4-6",
  "stop_reason": "end_turn",
  "stop_sequence": null,
  "stop_details": null,
  "usage": {
    "input_tokens": 17,
    "output_tokens": 12,
    "service_tier": "standard",
    "inference_geo": "global"
  }
}

Here is what each field means:

Field	Type	Description
`id`	string	Unique identifier for this message. Use for logging and debugging.
`type`	string	Always `message` for this endpoint.
`role`	string	Always `assistant` for responses.
`content`	array	Array of content blocks. Usually one text block.
`model`	string	The model that generated the response.
`stop_reason`	string	Why generation stopped. See stop reasons below.
`usage`	object	Token counts for input and output. Used for cost tracking.

Stop Reasons

The stop_reason field tells you why Claude stopped generating. Always check this field — a 200 OK does not guarantee a complete response.

Value	Meaning
`end_turn`	Claude finished naturally. Response is complete.
`max_tokens`	Generation stopped because `max_tokens` was reached. Response may be truncated.
`stop_sequence`	A custom stop sequence was triggered.
`tool_use`	Claude is requesting a tool call. See the Tool Use guide.

A 200 OK response with stop_reason: "max_tokens" and an empty content array means your max_tokens value is too low. Claude accepted the request but had no room to generate a response. Always set max_tokens to a reasonable value for your use case.

Required Request Fields

Every request to /v1/messages requires these three fields:

Field	Type	Description
`model`	string	The Claude model to use. See available models below.
`max_tokens`	integer	Maximum number of tokens to generate. Must be greater than `0`.
`messages`	array	The conversation history. Must contain at least one user message.

Available Models

Model	Description
`claude-sonnet-4-6`	Latest model. Best balance of speed and capability.
`claude-opus-4-6`	Most capable model. Best for complex tasks.
`claude-haiku-4-5`	Fastest and most compact model. Best for simple tasks.

Next Steps

Core Concepts

Understand how the messages array works and how context is managed

System Prompts

Learn how to shape Claude’s behavior for your use case

Streaming

Stream responses in real time for a better user experience

Tool Use

Connect Claude to real-time data and external services

Getting Started

Core Concepts

Guides

Prerequisites

Your First API Call

Using Postman

Using curl

Understanding the Response

Stop Reasons

Required Request Fields

Available Models

Next Steps

Core Concepts

System Prompts

Streaming

Tool Use

​Prerequisites

​Your First API Call

​Using Postman

​Using curl

​Understanding the Response

​Stop Reasons

​Required Request Fields

​Available Models

​Next Steps

Core Concepts

System Prompts

Streaming

Tool Use

Prerequisites

Your First API Call

Using Postman

Using curl

Understanding the Response

Stop Reasons

Required Request Fields

Available Models

Next Steps