How do I integrate image recognition onto an Assistant API?

I am building an analyzer for a client that analyzes trades based on images generated. I want it to use GPT-vision and I want to build it in playground with no code. is this possible? if not, how do i go about it?

It is no part of Assistants yet - but it’s promised to come.

Ok thanks, is there any external api i can call to basically do the job for me?