Function Calling via Chat Completions Tools — KiosAPI

Pass a tools array to POST /v1/chat/completions to enable function calling. The model returns structured JSON arguments for your functions.

You can enable any compatible chat model to call your own functions by including a tools array in your request to POST /v1/chat/completions. Instead of returning a plain text reply, the model will return structured JSON arguments matching the schema you define — it is your responsibility to execute the function and, if needed, send the result back in a follow-up message.

For a deeper conceptual overview of how function calling works, see the OpenAI Function Calling Guide.

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

The model to use for the request (e.g. gpt-4o).

messages · array · required

An array of message objects representing the conversation history so far.

tools · array · required

An array of tool definitions that the model may call. Each element describes one callable function.

Show tools[] properties

tools[].type · string · required

The type of tool. Must always be "function".

tools[].function.name · string · required

The name of the function. Allowed characters: a-z, A-Z, 0-9, underscores, and dashes. Maximum 64 characters.

tools[].function.description · string

A plain-language description of what the function does. Providing a clear description helps the model decide when to call it.

tools[].function.parameters · object · required

A JSON Schema object that describes the function’s parameters. Typically uses "type": "object" with a properties map and a required array.

tools[].function.strict · boolean

When set to true, the model will strictly adhere to the parameter schema you provide, refusing to pass any fields not defined in parameters.

Example Request

cURL

curl https://kiosapi.com/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer sk-xxx" \
  -d '{
    "model": "gpt-4o",
    "messages": [
      {
        "role": "user",
        "content": "What is the weather like in Paris today?"
      }
    ],
    "tools": [
      {
        "type": "function",
        "function": {
          "name": "get_weather",
          "description": "Get current temperature for a given location.",
          "parameters": {
            "type": "object",
            "properties": {
              "location": {
                "type": "string",
                "description": "City and country e.g. Paris, France"
              }
            },
            "required": ["location"],
            "additionalProperties": false
          },
          "strict": true
        }
      }
    ]
  }'

Python

import openai
import json

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

response = openai.chat.completions.create(
    model="gpt-4o",
    messages=[
        {"role": "user", "content": "What is the weather like in Paris today?"}
    ],
    tools=[
        {
            "type": "function",
            "function": {
                "name": "get_weather",
                "description": "Get current temperature for a given location.",
                "parameters": {
                    "type": "object",
                    "properties": {
                        "location": {
                            "type": "string",
                            "description": "City and country e.g. Paris, France"
                        }
                    },
                    "required": ["location"],
                    "additionalProperties": False
                },
                "strict": True
            }
        }
    ]
)

# The model returns a function call
tool_call = response.choices[0].message.tool_calls[0]
args = json.loads(tool_call.function.arguments)
print(args)  # {"location": "Paris, France"}

Response

When the model decides to call a function, the response contains a tool_calls array in the message object rather than a text content value.

{
  "id": "chatcmpl-Ax2bU1RFE8P0Y9uqKcErQNLcx4dDe",
  "object": "chat.completion",
  "created": 1738634724,
  "model": "gpt-4o-2024-08-06",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": null,
        "tool_calls": [
          {
            "id": "call_FezxjoWuDV1CL3dITVFOjUzK",
            "type": "function",
            "function": {
              "name": "get_weather",
              "arguments": "{\"location\":\"Paris, France\"}"
            }
          }
        ],
        "refusal": null
      },
      "logprobs": null,
      "finish_reason": "tool_calls"
    }
  ],
  "usage": {
    "prompt_tokens": 65,
    "completion_tokens": 16,
    "total_tokens": 81
  },
  "system_fingerprint": "fp_f3927aa00d"
}

Response Fields

choices[].message.tool_calls · array

Present when the model wants to invoke one or more functions. Each element contains an id, a type of "function", and a function object with name and arguments (a JSON-encoded string).

choices[].finish_reason · string

Set to "tool_calls" when the model is requesting a function invocation rather than producing a final text response.

When finish_reason is "tool_calls", you should parse the arguments string, execute the function on your end, and send the result back to the model in a new message with role: "tool". The model will then produce its final response based on that result.