Claude (Native Format) - PDF Analysis

Use Claude models via native Anthropic format with caching support

Claude (Native Format) - PDF Analysis

Claude’s native request format supports prompt caching, which can significantly reduce costs for repeated context.

The native format only lists commonly used parameters. For full parameter details, read the Claude official documentation.

Supported Models

Claude native format supports models with cld- or claude- prefixes, as well as kimi-k2-0711-preview, qwen3-coder-plus, and glm-4.5+.

Request

Headers

Header	Required	Description
Content-Type	Yes	application/json
x-api-key	Yes	sk-xxx
anthropic-version	Yes	2023-06-01

Request Body

Parameter	Type	Required	Description
model	string	Yes	Claude model name
messages	array	Yes	Array of message objects
messages[].role	string	Yes	Role: user or assistant
messages[].content	string/array	Yes	Text or multimodal content
max_tokens	number	No	Maximum tokens to generate
temperature	number	No	Sampling temperature (0-2)
top_p	number	No	Nucleus sampling
top_k	integer	No	Top-k sampling
stream	boolean	No	Enable streaming
tools	object	No	Function calling tools
thinking	object	No	Extended thinking config

Content Types for PDF/Image Analysis

Type	Sub-type	Fields	Description
image or document	base64	source.type: "base64", source.data, source.media_type	Base64-encoded file
image or document	url	source.type: "url", source.url	File URL
text	-	text	Text prompt

Cache Control

For files used across multiple messages, add cache_control to cache the file:

{
  "type": "document",
  "source": {
    "type": "base64",
    "data": "JVBERi0xLjQK...",
    "media_type": "application/pdf"
  },
  "cache_control": {
    "type": "ephemeral"
  }
}

Example Request - Chat

curl https://kiosapi.com/v1/messages \
  -H "Content-Type: application/json" \
  -H "x-api-key: sk-xxx" \
  -H "anthropic-version: 2023-06-01" \
  -d '{
    "model": "claude-sonnet-4-6",
    "messages": [
      {
        "role": "user",
        "content": "Hello, who are you?"
      }
    ],
    "max_tokens": 1688,
    "temperature": 0.5,
    "stream": false
  }'

Python Example - PDF Analysis with Caching

import anthropic

client = anthropic.Anthropic(
    api_key="sk-xxx",
    base_url="https://kiosapi.com"
)

response = client.messages.create(
    model="claude-sonnet-4-6",
    max_tokens=1000,
    messages=[
        {
            "role": "user",
            "content": [
                {
                    "type": "text",
                    "text": "Summarize this document"
                },
                {
                    "type": "document",
                    "source": {
                        "type": "url",
                        "url": "https://example.com/document.pdf"
                    },
                    "cache_control": {"type": "ephemeral"}
                }
            ]
        }
    ]
)

print(response.content[0].text)

Example - Extended Thinking

response = client.messages.create(
    model="claude-sonnet-5",
    max_tokens=8000,
    thinking={
        "type": "enabled",
        "budget_tokens": 4000
    },
    messages=[{"role": "user", "content": "Solve this complex problem..."}]
)

Response

{
  "id": "msg_014AoBefsejHUjbdRntn7euw",
  "type": "message",
  "role": "assistant",
  "content": [
    {
      "type": "text",
      "text": "Hello! I'm Claude, an AI assistant. How can I help you today?"
    }
  ],
  "model": "claude-sonnet-4-6",
  "stop_reason": "end_turn",
  "stop_sequence": null,
  "usage": {
    "input_tokens": 12,
    "output_tokens": 38
  }
}

Response Schema

Field	Type	Description
id	string	Message ID
type	string	Always message
role	string	Always assistant
content	array	Response content blocks
content[].type	string	text
content[].text	string	Response text
model	string	Model used
stop_reason	string	end_turn, max_tokens, tool_use
usage.input_tokens	integer	Input tokens
usage.output_tokens	integer	Output tokens