Claude API Examples

This page provides examples of using the Agentsflare Claude API to help you quickly integrate and use Claude AI services.

Basic Configuration

Before starting to use the API, please ensure you have obtained an API Key. If not, please refer to Create API Key.

Basic Information

API Base URL: https://api.agentsflare.com/anthropic/v1/messages
Authentication Method: API Key (x-api-key header)
Content Type: application/json

Request Examples

bash

curl --location --request POST 'https://api.agentsflare.com/anthropic/v1/messages' \
--header 'Content-Type: application/json' \
--header 'x-api-key: YOUR_API_KEY' \
--header 'anthropic-version: 2025-06-18' \
--data-raw '{
    "model": "claude-sonnet-5",
    "max_tokens": 1024,
    "messages": [
        {
            "role": "user",
            "content": "Hello, please introduce yourself"
        }
    ]
}'

python

import anthropic

client = anthropic.Anthropic(
    api_key="YOUR_API_KEY",
    base_url="https://api.agentsflare.com/anthropic"
)

message = client.messages.create(
    model="claude-sonnet-5",
    max_tokens=1024,
    messages=[
        {
            "role": "user",
            "content": "Hello, please introduce yourself"
        }
    ]
)

print(message.content[0].text)

python

import anthropic

client = anthropic.Anthropic(
    api_key="YOUR_API_KEY",
    base_url="https://api.agentsflare.com/anthropic"
)

message = client.messages.create(
    model="claude-sonnet-5",
    max_tokens=1024,
    messages=[
        {
            "role": "user",
            "content": "Hello, please introduce yourself"
        }
    ],
    stream=True
)

for event in message:
    if event.type == "content_block_delta":
        delta = event.delta
        if delta.type == "text_delta":
            print(delta.text, end="", flush=True)

javascript

import Anthropic from "@anthropic-ai/sdk";

const client = new Anthropic({
  apiKey: process.env.AGENTSFLARE_API_KEY,
  baseURL: "https://api.agentsflare.com/anthropic"
});

async function main() {
  try {
    const message = await client.messages.create({
      model: "claude-sonnet-5",
      max_tokens: 1024,
      messages: [
        {
          role: "user",
          content: "Hello, please introduce yourself"
        }
      ]
    });

    console.log(message.content[0].text);
  } catch (err) {
    console.error(err?.response?.data ?? err);
  }
}

main();

javascript

import Anthropic from "@anthropic-ai/sdk";

const client = new Anthropic({
  apiKey: process.env.AGENTSFLARE_API_KEY,
  baseURL: "https://api.agentsflare.com/anthropic"
});

async function main() {
  try {
    const stream = await client.messages.create({
      model: "claude-sonnet-5",
      max_tokens: 1024,
      messages: [
        {
          role: "user",
          content: "Hello, please introduce yourself"
        }
      ],
      stream: true
    });

    for await (const event of stream) {
      if (event.type === 'content_block_delta') {
        if (event.delta.type === 'text_delta') {
          process.stdout.write(event.delta.text);
        }
      }
    }
  } catch (err) {
    console.error(err?.response?.data ?? err);
  }
}

main();

javascript

const Anthropic = require("@anthropic-ai/sdk");

const client = new Anthropic({
  apiKey: process.env.AGENTSFLARE_API_KEY,
  baseURL: "https://api.agentsflare.com/anthropic"
});

async function main() {
  try {
    const message = await client.messages.create({
      model: "claude-sonnet-5",
      max_tokens: 1024,
      messages: [
        {
          role: "user",
          content: "Hello, please introduce yourself"
        }
      ]
    });

    console.log(message.content[0].text);
  } catch (err) {
    console.error(err?.response?.data ?? err);
  }
}

main();

package main

import (
	"context"
	"fmt"
	"log"
	"os"

	"github.com/anthropics/anthropic-sdk-go"
	"github.com/anthropics/anthropic-sdk-go/option"
)

func main() {
	apiKey := os.Getenv("AGENTSFLARE_API_KEY")
	if apiKey == "" {
		log.Fatal("missing env AGENTSFLARE_API_KEY")
	}

	client := anthropic.NewClient(
		option.WithAPIKey(apiKey),
		option.WithBaseURL("https://api.agentsflare.com/anthropic"),
	)

	ctx := context.Background()

	message, err := client.Messages.New(ctx, anthropic.MessageNewParams{
		Model:     anthropic.F("claude-sonnet-5"),
		MaxTokens: anthropic.F(int64(1024)),
		Messages: anthropic.F([]anthropic.MessageParam{
			anthropic.NewUserMessage(anthropic.NewTextBlock("Hello, please introduce yourself")),
		}),
	})

	if err != nil {
		log.Fatalf("message creation failed: %v", err)
	}

	fmt.Println(message.Content[0].Text)
}

Response Examples

Non-streaming Response

json

{
  "id": "msg_01XFDUDYJgAACzvnptvVoYEL",
  "type": "message",
  "role": "assistant",
  "content": [
    {
      "type": "text",
      "text": "Hello! I'm Claude, an AI assistant created by Anthropic. I'm designed to be helpful, harmless, and honest. I can assist you with a wide variety of tasks including writing, analysis, programming, learning, and more. How can I help you today?"
    }
  ],
  "model": "claude-sonnet-5",
  "stop_reason": "end_turn",
  "stop_sequence": null,
  "usage": {
    "input_tokens": 15,
    "output_tokens": 65
  }
}

Streaming Response

json

event: message_start
data: {"type":"message_start","message":{"id":"msg_123","type":"message","role":"assistant","content":[],"model":"claude-sonnet-5","stop_reason":null,"usage":{"input_tokens":15,"output_tokens":0}}}

event: content_block_start
data: {"type":"content_block_start","index":0,"content_block":{"type":"text","text":""}}

event: content_block_delta
data: {"type":"content_block_delta","index":0,"delta":{"type":"text_delta","text":"Hello"}}

event: content_block_delta
data: {"type":"content_block_delta","index":0,"delta":{"type":"text_delta","text":"!"}}

event: content_block_stop
data: {"type":"content_block_stop","index":0}

event: message_delta
data: {"type":"message_delta","delta":{"stop_reason":"end_turn","stop_sequence":null},"usage":{"output_tokens":65}}

event: message_stop
data: {"type":"message_stop"}

Request Parameters

Parameter	Type	Required	Description
model	string	Yes	Model name, e.g., `claude-sonnet-5`
messages	array	Yes	Array of messages with role and content
max_tokens	integer	Yes	Maximum tokens to generate
stream	boolean	No	Enable streaming response, default false
system	string	No	System prompt

Note: On Claude Opus 4.7 and newer model families, setting temperature, top_p, or top_k to non-default values can return HTTP 400. Omit these parameters unless the target model explicitly supports them.

Features

Streaming Output

Claude API supports streaming output (SSE) by setting stream: true. Streaming responses allow real-time content generation, providing a better user experience.

System Prompts

Claude supports system prompts via the system parameter to define the assistant's role and behavior.

Structured Outputs

Use output_config.format when you need valid JSON that conforms to a JSON Schema. The format value must be an object with type: "json_schema" and a nested schema; it is not the string "json".

bash

curl -s https://api.agentsflare.com/anthropic/v1/messages \
  -H "content-type: application/json" \
  -H "x-api-key: YOUR_API_KEY" \
  -H "anthropic-version: 2025-06-18" \
  -d '{
    "model": "claude-sonnet-5",
    "max_tokens": 1024,
    "messages": [
      {"role": "user", "content": "Return a summary and key points."}
    ],
    "output_config": {
      "format": {
        "type": "json_schema",
        "schema": {
          "type": "object",
          "properties": {
            "summary": {"type": "string"},
            "key_points": {"type": "array", "items": {"type": "string"}}
          },
          "required": ["summary", "key_points"],
          "additionalProperties": false
        }
      }
    }
  }'

Multi-turn Conversations

Include previous user and assistant messages in the messages array:

python

messages = [
    {"role": "user", "content": "What is machine learning?"},
    {"role": "assistant", "content": "Machine learning is a branch of AI..."},
    {"role": "user", "content": "What are its applications?"},
]

message = client.messages.create(
    model="claude-sonnet-5",
    max_tokens=1024,
    messages=messages,
)

Claude 4.7+ Migration and Advanced Features

The following compatibility notes apply when migrating from older Claude models, especially from Opus 4.6 to Opus 4.7 or later models.

Change	Compatibility impact	Recommended action
Sampling parameters	Non-default `temperature`, `top_p`, or `top_k` values can return HTTP 400 on newer model families	Omit these fields unless the target model explicitly supports them
Thinking mode	`thinking.type: "enabled"` is deprecated on Claude 4.6 and rejected on Claude 4.7+	Use `thinking.type: "adaptive"` and control depth with `output_config.effort`
Assistant prefill	A final `assistant` prefill is unsupported starting with Claude 4.6 models	Put constraints in `system`, or use structured outputs for schema-constrained JSON
Thinking visibility	Thinking text is omitted by default	Set `display: "summarized"` when a readable thinking summary is required

Adaptive Thinking Example

bash

curl -s https://api.agentsflare.com/anthropic/v1/messages \
  -H "content-type: application/json" \
  -H "x-api-key: YOUR_API_KEY" \
  -H "anthropic-version: 2025-06-18" \
  -d '{
    "model": "claude-opus-4-7",
    "max_tokens": 16000,
    "messages": [
      {"role": "user", "content": "Analyze the trade-offs of this architecture."}
    ],
    "thinking": {
      "type": "adaptive",
      "display": "summarized"
    },
    "output_config": {
      "effort": "high"
    }
  }'

OpenAI SDK compatibility

When calling Claude through the OpenAI-compatible /v1 endpoint, pass thinking and output_config through extra_body. See Claude Reasoning for a complete example.

Prompt Caching

Claude API supports Prompt Caching, allowing you to cache large context blocks in system or messages (such as project specifications, reference documents, etc.). Subsequent requests can reuse the cached content, significantly reducing token consumption and response latency.

Limitations

Item	Limit
Claude Fable 5.1 / Opus 5 / Fable 5 / Mythos 5	512 tokens
Claude Opus 4.7	2,048 tokens
Claude Opus 4.6 / Opus 4.5 / Haiku 4.5	4,096 tokens
Claude Opus 4.8 / Sonnet 5 / Sonnet 4.6 / Sonnet 4.5	1,024 tokens
Default cache TTL	300 seconds (5 minutes); a 1-hour TTL is available only where supported

Prompts shorter than the applicable minimum are processed normally but are not cached, and no error is returned.

Usage

Use the array format for system and add the cache_control field to the content block you want to cache.

Recommendation

Keep stable content before the cache breakpoint and move dynamic values, timestamps, and request-specific text after it.

bash

curl -s https://api.agentsflare.com/anthropic/v1/messages \
  -H "content-type: application/json" \
  -H "x-api-key: YOUR_API_KEY" \
  -H "anthropic-version: 2025-06-18" \
  -d '{
    "model": "claude-sonnet-5",
    "max_tokens": 1024,
    "system": [
      {
        "type": "text",
        "text": "[Project: project-alpha]\n\nPlace your large context here, such as project specs, technical docs, coding standards...",
        "cache_control": {"type": "ephemeral"}
      }
    ],
    "messages": [
      {
        "role": "user",
        "content": "Based on the above specs, help me review this code."
      }
    ]
  }'

Response Examples

First request (cache creation):

json

{
  "usage": {
    "input_tokens": 29,
    "cache_creation_input_tokens": 2480,  
    "cache_read_input_tokens": 0,         
    "output_tokens": 655
  }
}

Subsequent request (cache hit):

json

{
  "usage": {
    "input_tokens": 33,
    "cache_creation_input_tokens": 0,     
    "cache_read_input_tokens": 2480,      
    "output_tokens": 1024
  }
}

Response Field Descriptions

Field	Description
`cache_creation_input_tokens`	Tokens newly cached in this request. A value greater than 0 indicates cache was created successfully
`cache_read_input_tokens`	Tokens read from cache in this request. A value greater than 0 indicates a successful cache hit
`input_tokens`	Regular (non-cached) input tokens, such as user messages in `messages`

Determining Cache Status

cache_creation_input_tokens > 0: Cache created successfully — the content has been cached and subsequent identical requests will hit the cache
cache_read_input_tokens > 0: Cache hit successfully — this request reused previously cached content, saving token consumption
Both are 0: Cache did not take effect — the content may be below the model-specific minimum, the prefix may have changed, or the cache may have expired

Cache Hit Rules and Troubleshooting

Cache matching is prefix-based. A byte-level change before a breakpoint invalidates that breakpoint and all content after it.
The cache hierarchy follows tools → system → messages; changing tools or switching models invalidates the affected prefix.
A request can contain up to 4 explicit cache breakpoints.
Avoid timestamps, random IDs, unstable JSON key ordering, or other dynamic content before a breakpoint.
Verify behavior with usage.cache_creation_input_tokens and usage.cache_read_input_tokens instead of assuming a cache hit.

Important Notes

API Key Security: Do not hardcode API Keys in your code, use environment variables
Request Rate: Please comply with API call rate limits
Error Handling: Implement comprehensive error handling mechanisms
Token Limits: Be aware of different models' context window limits

Supported Models

The following models are available through this interface (sorted by recommendation):

claude-fable-5-1 New
claude-opus-5 New
claude-mythos-5
claude-fable-5
claude-opus-4-8
claude-opus-4-7
claude-opus-4-6
claude-opus-4-5-20251101
claude-opus-4-1-20250805 Retired
claude-sonnet-5
claude-sonnet-4-6
claude-sonnet-4-5-20250929
claude-haiku-4-5-20251001
claude-sonnet-4-20250514 Retired
claude-opus-4-20250514 Retired
claude-3-haiku-20240307 Retired

💡 Tip

The model field in the request example can be replaced with any model name above.

Claude API Examples ​

Basic Configuration ​

Basic Information ​

Request Examples ​

Response Examples ​

Non-streaming Response ​

Streaming Response ​

Request Parameters ​

Features ​

Streaming Output ​

System Prompts ​

Structured Outputs ​

Multi-turn Conversations ​

Claude 4.7+ Migration and Advanced Features ​

Adaptive Thinking Example ​

Prompt Caching ​

Limitations ​

Usage ​

Response Examples ​

Response Field Descriptions ​

Cache Hit Rules and Troubleshooting ​

Important Notes ​

Related Links ​

Supported Models ​

Claude API Examples

Basic Configuration

Basic Information

Request Examples

Response Examples

Non-streaming Response

Streaming Response

Request Parameters

Features

Streaming Output

System Prompts

Structured Outputs

Multi-turn Conversations

Claude 4.7+ Migration and Advanced Features

Adaptive Thinking Example

Prompt Caching

Limitations

Usage

Response Examples

Response Field Descriptions

Cache Hit Rules and Troubleshooting

Important Notes

Related Links

Supported Models