Skip to main content

Overview

Background​

TaskingAI brings together the APIs from many different model providers into one unified platform. When you use TaskingAI's API to interact with models or assistants, you get responses in a consistent format defined by TaskingAI even if the model providers differ from each other.

However, as there are many existing projects built with OpenAI's API or OpenAI-compatible APIs, we have created an OpenAI-compatible API to make it easier for you to migrate your projects to TaskingAI with minimum changes in code.

How to Use​

Normal TaskingAI API calls are made to https://api.tasking.ai/, and the response will be of TaskingAI's format. To get OpenAI-compatible responses, all you need to do is change the base URL to https://oapi.tasking.ai/.

All endpoints of the OpenAI-compatible API are the same as the TaskingAI API. You can use the same endpoints and parameters as you would with the TaskingAI API.

Example​

Here is an example of how to use the OpenAI python SDK to interact with TaskingAI's service through the OpenAI-compatible API:

from openai import OpenAI

client = OpenAI(
    api_key="YOUR_TASKINGAI_API_KEY",
    base_url="https://oapi.tasking.ai/v1",
)

# If you prefer to directly interact with your TaskingAI models, replace with your model ID.
MODEL_ID = "YOUR_TASKINGAI_ASSISTANT_ID"

response = client.chat.completions.create(
        model=MODEL_ID,
        messages=[
            {"role": "user", "content": "Hello, how are you?"},
        ]
    )

print(response)

With the OpenAI-compatible API, you may utilize TaskingAI's service with almost all existing OpenAI-compatible SDKs and libraries without any changes in your code.