Making Requests

SynStar AI provides OpenAI-compatible API endpoints for common chat, vision, and multimodal workflows.

This page explains the basic structure of an API request.

Request Structure

A typical API request includes:

1. An API endpoint
2. Request headers
3. A JSON request body
4. A selected model ID
5. Input messages or content

For chat completions, the endpoint is:

POST https://www.kkiai.com/v1/chat/completions

HTTP Method

Most generation requests use the POST method.

POST /v1/chat/completions

Use POST when sending messages, prompts, images, parameters, or other request body data to a model.

Headers

Each request should include the following headers:

Content-Type: application/json
Authorization: Bearer YOUR_API_KEY

Header	Required	Description
Content-Type	Yes	Use application/json for JSON request bodies
Authorization	Yes	Your API key in Bearer format

Request Body

For a basic chat completion request, the body usually includes:

Parameter	Type	Required	Description
model	string	Yes	The model ID you want to use
messages	array	Yes	The conversation messages
temperature	number	No	Controls randomness
stream	boolean	No	Whether to return the response as a stream
max_tokens	integer	No	Maximum number of tokens to generate

Messages

The messages field is an array of message objects.

Each message usually contains:

Field	Type	Description
role	string	The role of the message sender
content	string or array	The message content

Common roles include:

system
user
assistant

Example:

[
  {
    "role": "system",
    "content": "You are a helpful assistant."
  },
  {
    "role": "user",
    "content": "Explain what an API gateway is."
  }
]

Example Request

curl https://www.kkiai.com/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{
    "model": "gpt-4o",
    "messages": [
      {
        "role": "system",
        "content": "You are a helpful assistant."
      },
      {
        "role": "user",
        "content": "Say this is a test!"
      }
    ],
    "temperature": 0.7
  }'

Example Response

A successful response will look similar to this:

{
  "id": "chatcmpl-abc123",
  "object": "chat.completion",
  "created": 1677858242,
  "model": "gpt-4o",
  "usage": {
    "prompt_tokens": 13,
    "completion_tokens": 7,
    "total_tokens": 20
  },
  "choices": [
    {
      "message": {
        "role": "assistant",
        "content": "This is a test!"
      },
      "finish_reason": "stop",
      "index": 0
    }
  ]
}

Response Fields

Field	Description
id	Unique ID of the response
object	Response object type
created	Unix timestamp of when the response was created
model	Model used for the request
usage	Token usage information
choices	Generated response choices

The generated text is usually located at:

choices[0].message.content

Streaming Requests

To receive output incrementally, set stream to true.

{
  "model": "gpt-4o",
  "messages": [
    {
      "role": "user",
      "content": "Write a short introduction to AI APIs."
    }
  ],
  "stream": true
}

Streaming is useful for chat interfaces because the response can be displayed as it is generated.

Choosing a Model

Set the model parameter to a supported model ID.

Example:

{
  "model": "gpt-4o"
}

You can find supported model IDs in the model list or model marketplace.

Use the model name exactly as shown in the documentation or dashboard.

Common Request Issues

Missing API Key

Make sure the request includes:

Authorization: Bearer YOUR_API_KEY

Incorrect Base URL

For most SDK integrations, use:

https://www.kkiai.com/v1

For direct chat completion requests, use:

https://www.kkiai.com/v1/chat/completions

Invalid Model ID

Check that the model ID is spelled correctly and available in your account.

Insufficient Balance

If the request fails due to balance or quota, check your account balance, token quota, and token group settings in the dashboard.