Changing the realtime-console example to 4o-mini-realtime-preview

ChatGPT
1

I’ve been checking out the realtime-console example given in the realtime api documentation. But my previous experience is with Python, so I’m having a difficult time tracking down how and where I can change the model being used from 4o-preview to 4o-mini-preview. ConsolePage.tsx has some portions that look promising, but no matter what I modify, I can’t seem to make the model change.

I’ve looked through the realtime-api-beta library to no avail. I’m hoping that someone with the language-specific experience I lack will be able to answer this easily. Thanks for reading, whether you’re able to help or not.

2

I’m having the same problem. Have you figured it out yet?

3

No, and to make matters worse the repo was update yesterday to a completely different implementation that doesn’t even load. I’m jealous of webdevs, but I might have to end up using the less pretty Python example I should at least know how to change variables in.

4

Work around solution I found is:

Replace: await client.connect()

With: await client.realtime.connect({ model: “YOUR MODEL HERE”});

5

this didnt work for me…anyone figured it out yet?

6

pr 103 - thats the fix you can finally set custom models!

1 Like
7

Can you please send the link for pull request 103? I don’t see any PR with such a number in the repo

8
2 Likes
9

This topic was automatically closed 2 days after the last reply. New replies are no longer allowed.