API Performance

Community Discussion
1

I am using Bubble for frontend development.

Using the Mindstudio embed functionality, the outputs are being streamed as they are received by the model being used.

In Bubble, the outputs appear to be batched, which requires the user to wait for the entire response to be receive impacting the user experience.

What can be done to address this?

2

You can try handling the stream outside of Bubble using custom JavaScript, then update Bubble elements in real time through the toolbox plugin or JavaScript to Bubble event. It’s a bit of extra setup but it can give you that streaming effect like in chat apps.