Anthropic Knowledge Hub

SOP: Build with the Messages API ​

Updated: 2026-05-26

Make effective use of the Messages API: streaming, multi-turn conversations, and vision.

Prerequisites ​

  • API key configured
  • Basic understanding of the Messages API

Steps ​

Step 1: Basic Request ​

bash
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"}
    ]
  }'

Step 2: Enable Streaming ​

Add "stream": true to receive server-sent events:

bash
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,
    "stream": true,
    "messages": [
      {"role": "user", "content": "Write a short poem"}
    ]
  }'

Handle events: message_start, content_block_delta, message_delta, message_stop.

Step 3: Multi-turn Conversation ​

Maintain conversation history by including all prior messages:

python
messages = [
    {"role": "user", "content": "What is Python?"},
    {"role": "assistant", "content": "Python is a programming language..."},
    {"role": "user", "content": "Show me an example"}
]

Step 4: Vision with Images ​

Include images as base64-encoded data or image URLs:

python
messages = [
    {"role": "user", "content": [
        {"type": "image", "source": {
            "type": "base64",
            "media_type": "image/jpeg",
            "data": "<base64-encoded-image>"
        }},
        {"type": "text", "text": "Describe this image"}
    ]}
]

Step 5: System Prompt ​

Set system instructions for role and behavior:

python
response = client.messages.create(
    model="claude-sonnet-4-6",
    max_tokens=1024,
    system="You are a helpful code reviewer. Review code for correctness, efficiency, and readability.",
    messages=[{"role": "user", "content": "def foo(x): return x+1"}]
)

Verification Checklist ​

  • [ ] Basic request returns assistant text response
  • [ ] Streaming events include message_start and message_stop
  • [ ] Multi-turn maintains context across exchanges
  • [ ] Image input is accepted and described
  • [ ] System prompt influences assistant behavior

See Also ​