This integration is only for Keywords AI LLM proxy user. If you only want to log LLM calls and responses, you can use the Logging API to log LLM calls.

This integration works exclusively with OpenAI models and cannot be used with models from other providers.

Responses API is OpenAI’s most advanced interface for generating model responses. Supports text and image inputs, and text outputs. Create stateful interactions with the model, using the output of previous responses as input.

Extend the model’s capabilities with built-in tools for file search, web search, computer use, and more. Allow the model access to external systems and data using function calling.

This guide will show you how to integrate the Responses API with Keywords AI so you can log LLM calls from Responses API.

​
Quickstart

​
Prerequisites

To use Keywords AI gateway, you need to:

  1. Create an API key in Keywords AI platform.
  2. Add your OpenAI API key in Providers to call models from OpenAI.

​
Create a model response / Text input example

You can use Keywords AI gateway and get observability by changing the base url and adding a custom header.

from openai import OpenAI
from base64 import b64encode
import json

client = OpenAI(
    base_url="https://api.keywordsai.co/api/chat/completions",
    api_key="Your_Keywords_AI_API_Key",
)

keywordsai_params = {
    "metadata": {
        "paid_user": "true",
    }
    # Other keywordsai params
}

keywordsai_params_header = {
    "X-Data-Keywordsai-Params": b64encode(json.dumps(keywordsai_params).encode()).decode(),
}

def test_text_input():
    response = client.responses.create(
    model="gpt-4o",
    input="Tell me a three sentence bedtime story about a unicorn.",
    extra_headers=keywordsai_params_header,
    )

    print(response)

if __name__ == "__main__":
    test_text_input()

​
File Search example

from openai import OpenAI
from base64 import b64encode
import json

client = OpenAI(
    base_url="https://api.keywordsai.co/api/chat/completions",
    api_key="Your_Keywords_AI_API_Key",
)

keywordsai_params = {
    "metadata": {
        "paid_user": "true",
    }
    # Other keywordsai params
}


keywordsai_params_header = {
    "X-Data-Keywordsai-Params": b64encode(json.dumps(keywordsai_params).encode()).decode(),
}

def test_file_search():
    response = client.responses.create(
        model="gpt-4o",
        tools=[
            {
                "type": "file_search",
                "vector_store_ids": ["vs_67d3bdd0c8888191adfa890a9e829480"],
                "max_num_results": 20,
            }
        ],
        input="What are the attributes of an ancient brown dragon?",
        extra_headers=keywordsai_params_header,
    )


if __name__ == "__main__":
    test_file_search()

​
Reasoning example

from openai import OpenAI
from base64 import b64encode
import json

client = OpenAI(
    base_url="https://api.keywordsai.co/api/chat/completions",
    api_key="Your_Keywords_AI_API_Key",
)
keywordsai_params = {
    "metadata": {
        "paid_user": "true",
    }
    # Other keywordsai params
}

keywordsai_params_header = {
    "X-Data-Keywordsai-Params": b64encode(json.dumps(keywordsai_params).encode()).decode(),
}
def test_file_search():
  response = client.responses.create(
    model="o3-mini",
    input="How much wood would a woodchuck chuck?",
    reasoning={
        "effort": "high"
    },
    extra_headers=keywordsai_params_header,
  )

  print(response)


if __name__ == "__main__":
    test_file_search()

​
Streaming example

from openai import OpenAI
from base64 import b64encode
import json

client = OpenAI(
    base_url="https://api.keywordsai.co/api/chat/completions",
    api_key="Your_Keywords_AI_API_Key",
)

keywordsai_params = {
    "metadata": {
        "paid_user": "true",
    }
    # Other keywordsai params
}

keywordsai_params_header = {
    "X-Data-Keywordsai-Params": b64encode(json.dumps(keywordsai_params).encode()).decode(),
}

def test_file_search():
    response = client.responses.create(
        model="gpt-4o",
        instructions="You are a helpful assistant.",
        input="Hello!",
        stream=True,
        extra_headers=keywordsai_params_header,
    )

    for chunk in response:
        print(chunk)


if __name__ == "__main__":
    test_file_search()

​
Functions example

from openai import OpenAI
from base64 import b64encode
import json

client = OpenAI(
    base_url="https://api.keywordsai.co/api/chat/completions",
    api_key="Your_Keywords_AI_API_Key",
)

keywordsai_params = {
    "metadata": {
        "paid_user": "true",
    }
    # Other keywordsai params
}

keywordsai_params_header = {
    "X-Data-Keywordsai-Params": b64encode(json.dumps(keywordsai_params).encode()).decode(),
}

def test_file_search():
  tools = [
        {
            "type": "function",
            "name": "get_current_weather",
            "description": "Get the current weather in a given location",
            "parameters": {
            "type": "object",
            "properties": {
              "location": {
                  "type": "string",
                  "description": "The city and state, e.g. San Francisco, CA",
              },
              "unit": {"type": "string", "enum": ["celsius", "fahrenheit"]},
            },
            "required": ["location", "unit"],
        }
        }
    ]

  response = client.responses.create(
        model="gpt-4o",
        tools=tools,
        input="What is the weather like in Boston today?",
        tool_choice="auto",
        extra_headers=keywordsai_params_header,
    )

  print(response)


if __name__ == "__main__":
    test_file_search()

​
Web Search example

from openai import OpenAI
from base64 import b64encode
import json

client = OpenAI(
    base_url="https://api.keywordsai.co/api/chat/completions",
    api_key="Your_Keywords_AI_API_Key",
)   

keywordsai_params = {
    "metadata": {
        "paid_user": "true",
    }
    # Other keywordsai params
}

keywordsai_params_header = {
    "X-Data-Keywordsai-Params": b64encode(json.dumps(keywordsai_params).encode()).decode(),
}

def test_web_search():

    response = client.responses.create(
        model="o3-mini",
        input="How much wood would a woodchuck chuck?",
        reasoning={
            "effort": "high"
        },
        extra_headers=keywordsai_params_header,
    )

    print(response.model_dump())

if __name__ == "__main__":
    test_web_search()

Was this page helpful?

Suggest editsRaise issue