Native Claude Format

Introduction

Claude's native message interface, suitable for native Anthropic clients such as Claude Code. This interface complies with Anthropic's API specifications and provides full feature support for Claude models, including advanced capabilities like Extended Thinking and tool calling.

You can test different scenarios by switching the Body example below or by setting the stream field to true.

All requests share the following headers:

x-api-key: sk-xxx (or use Authorization)

anthropic-version: 2023-06-01

content-type: application/json

When stream=true, the response will be an SSE stream (content-type: text/event-stream), and the event types follow the native Anthropic specification (message_start / content_block_start / content_block_delta / content_block_stop / message_delta / message_stop).

Prompt Caching

Prompt Caching allows you to cache frequently used context, significantly reducing costs and improving response times. There are two ways to enable it, which can be used individually or in combination:

Approach	Usage	Recommended Use Case
Auto-Caching	Place `cache_control` at the top level of the request body	Multi-turn conversations. The breakpoint automatically lands on the last cacheable block and moves forward with the conversation, eliminating the need to manually maintain tags.
Explicit Breakpoints	Add `cache_control` markers directly to content blocks in `system` or `messages`	Precise cache control. Ideal when you need to cache specific content boundaries, such as only caching the system prompt or a specific long document.

curl --location 'https://api.aboai.ai/v1/messages' \ --header 'Content-Type: application/json' \ --data '{ "model": "claude-sonnet-4.5", "max_tokens": 1024, "messages": [ { "role": "user", "content": "Hello, nice to meet you" } ] }'

{ "model": "claude-sonnet-4-5-20250929", "id": "msg_bdrk_01DJBYweRg351XHSgq6JcTsE", "type": "message", "role": "assistant", "content": [ { "type": "text", "text": "Hello! Nice to meet you too! How are you doing today? Is there anything I can help you with?" } ], "stop_reason": "end_turn", "usage": { "input_tokens": 13, "cache_creation_input_tokens": 0, "cache_read_input_tokens": 0, "cache_creation": { "ephemeral_5m_input_tokens": 0, "ephemeral_1h_input_tokens": 0 }, "output_tokens": 26 } }

Native Claude Format

Introduction

Prompt Caching

Request

Responses

Native Claude Format

Introduction#

Prompt Caching#

Request

Responses

Introduction

Prompt Caching