Thank you for your kind words and for reaching out with your feature request.

Could you please specify which API endpoint you'd like us to offer in a format compatible with the OpenAI API standards? We will investigate the feasibility of this integration and consider it for future updates :-)

For reference, you can check TaskingAI API endpoints at https://docs.tasking.ai/api/

Thanks for your reply！！！
The most commonly used OpenAI API endpoint is: /v1/chat/completions
Using this endpoint, you can access the established assistant in a streaming output manner.
You may need to bind the API key and assistant together, so that once authentication is complete, you can directly access the corresponding assistant.

@TomZhou2024 OpenAI-compatible chat completion API is now available in the latest release, v0.3.0. Please try it out!

@jameszyao @SimsonW @XiaoZhou2024 Hello, thanks for the feature. How can I make the API "OpenAI-compatible" when I self-host TaskingAI?

@jameszyao @SimsonW @XiaoZhou2024 Hello, thanks for the feature. How can I make the API "OpenAI-compatible" when I self-host TaskingAI?

You may find the instructions by: go to assistant or model management page -> locate your target model or assistant -> click the three dots at the right of that row -> click 'View Code'. You will find something similar to the following screenshot:

This includes all the possible ways to use your assistants and models out of the UI console.

Or If you are using Python, please follow the instructions here: https://www.tasking.ai/examples/access-taskingai-agents-through-openai-compatible-apis

This post was for using cloud platform. And since you are self-hosting, simply replace base_url to http://127.0.0.1:8080/v1 (assuming you did not change the service port) and it should work.