Send model_name or checkpoint_name to "/sdapi/v1/txt2img" API for generate images with different model.

[hosein-moayedi]

I am running Automatic1111 as a backend service and using those API endpoints. (--api)
I saved some different trained models (checkpoints .ckpt) files on my Linux server and I want to use each one for generating new images using text2image for each request from users.
Is there any way to do that?

Because in this time I think, in the first step we should change the selected checkpoint to the target model and after that, we should start to generate a new image by text2image.

However, if user 1 selects model 1 and he is generating a new image from that model and at the same time, user 2 wants select model 2 and wants to generate a new image from model 2, one of that users will use the wrong model! (ex: the image of user 1 generated from model 2 because the last model that was selected was model 2)

How should I handle it?

Is there any way to send the model name or checkpoint name for each request to this endpoint: "/sdapi/v1/txt2img"?