Claude via OpenAI Format — PDF Analysis and Thinking

Use Claude models at POST /v1/chat/completions with any OpenAI-compatible SDK. Supports PDF analysis, extended thinking, image inputs, and streaming.

KiosAPI lets you call Claude models using the standard OpenAI request format — the same endpoint, headers, and SDK you already use for GPT models. You send requests to /v1/chat/completions with your Claude model name, and PDFs and other files are supported by passing structured objects in the content array of each message.

If you need prompt caching, use the Claude native format instead — the OpenAI-compatible format does not support PDF caching. See the Claude (Native) page for details.

Request

POST https://kiosapi.com/v1/chat/completions

Headers

Header	Required	Description
Content-Type	Yes	application/json
Authorization	Yes	Bearer sk-xxx

Request Body

model · string · required

Claude model name (e.g. claude-sonnet-4-6).

messages · array · required

Array of message objects forming the conversation.

Show messages properties

messages[].role · string · required

Role of the message author: user or assistant.

messages[].content · string | array · required

The message content. Pass a plain string for text-only messages, or an array of content objects for multimodal input such as PDFs and images.

max_tokens · number

Maximum number of tokens to generate in the response.

temperature · number

Sampling temperature between 0 and 2.

stream · boolean

Set to true to stream the response using server-sent events.

thinking · object

Enable extended thinking. Available on claude-3-7-sonnet only.

Show thinking properties

thinking.type · string

Must be "enabled" to activate extended thinking.

thinking.budget_tokens · integer

Maximum tokens the model may use for internal reasoning. Minimum 1024; must be less than max_tokens.

Content Types for PDF and File Analysis

When passing content as an array, each element uses one of the following types:

Type	Fields	Description
text	text	Plain text prompt
file	file.filename, file.file_data	PDF file passed as a URL or base64 data
file_url	file_url.url	File URL (supports images and PDFs)
image_url	image_url.url	Image URL (base64 data URI or web URL)

For PDF files, Claude recommends base64 encoding for faster results. Use the format data:application/pdf;base64,JVBERi0xLjQK....

Examples

cURL (Chat)

curl https://kiosapi.com/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer sk-xxx" \
  -d '{
    "model": "claude-sonnet-4-6",
    "messages": [
      {
        "role": "user",
        "content": "Hello, who are you?"
      }
    ],
    "max_tokens": 1688,
    "temperature": 0.5,
    "stream": false
  }'

cURL (PDF)

curl https://kiosapi.com/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer sk-xxx" \
  -d '{
    "model": "claude-sonnet-4-6",
    "messages": [
      {
        "role": "user",
        "content": [
          {
            "type": "text",
            "text": "Summarize this document"
          },
          {
            "type": "file_url",
            "file_url": {
              "url": "https://example.com/document.pdf"
            }
          }
        ]
      }
    ],
    "max_tokens": 1000,
    "stream": false
  }'

Python (PDF Analysis)

import openai

openai.api_key = "sk-xxx"
openai.base_url = "https://kiosapi.com/v1/"

response = openai.chat.completions.create(
    model="claude-sonnet-4-6",
    messages=[
        {
            "role": "user",
            "content": [
                {"type": "text", "text": "Summarize this document"},
                {
                    "type": "file",
                    "file": {
                        "filename": "document.pdf",
                        "file_data": "https://example.com/document.pdf"
                    }
                }
            ]
        }
    ],
    max_tokens=1000,
    stream=False
)

print(response.choices[0].message.content)

Python (Extended Thinking)

response = openai.chat.completions.create(
    model="claude-3-7-sonnet-20250219",
    messages=[{"role": "user", "content": "Solve this complex problem..."}],
    max_tokens=8000,
    thinking={
        "type": "enabled",
        "budget_tokens": 4000
    }
)

Response

{
  "id": "chatcmpl-xxx",
  "object": "chat.completion",
  "created": 1724972230,
  "model": "claude-sonnet-4-6",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "Hello! I'm Claude, an AI assistant..."
      },
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 12,
    "completion_tokens": 38,
    "total_tokens": 50
  }
}