This is a demo of our brand new assistant settings.
00:04
So on the homepage here, I'm going to click on the little gear icon on this prompt box.
00:09
When I open it, you're going to see the new assistant settings.
00:13
Now, we took all of the different form fields and options that were there previously and we just reorganized them, put some better names around them, made it clearer and easier to find them and understand what they're doing.
00:24
So we took them and put them into standard settings, which is really most used.
00:28
General settings consulted evidence, the reference preferences, the reference types, and then the output type.
00:36
Now, part of why we did this was obviously just to make it better and easier for you to find the right settings.
00:41
But we also wanted to set ourselves up to continue to expand on this as we add new data sources, as we add new output types, we want to have a natural place to let those be displayed so that you can find them moving forward in the standard settings.
00:57
I do want to call out that we have a brand new model selector.
01:00
So we spent some time thinking about how we can make this easier to understand the difference between selecting something like GPT4O mini as our default to something like GPT5 mini.
01:12
Now, most models really look at things on two vectors of speed and intelligence.
01:17
So if it's going to be faster, it's probably gonna have a little less intelligence in it.
01:22
If it's gonna take longer and have lower speed, then it's probably gonna be a thinking model or have higher intelligence to it.
01:28
You need to weigh this based off of the type of question you're asking.
01:31
And we wanna continue to improve this because we don't even want you to have to make that decision.
01:35
So we think there's a lot of benefit in setting this up because we can improve it over time.
01:39
So you can see here we have this, little rating of speed and intelligence listing next to each of the models to help you pick what you're looking for.
01:47
And then we also released two brand new models, GPT 5.2 and Claude 4.5 Opus.
01:55
These are brand new models just released and you could see them relative to our speed and intelligence ratings here.
02:01
I also want to call out, that these are relative to the models that we provide and you have access to with insight.
02:07
These are not relative to all models that exist.
02:10
The speed and intelligence would be different if you look at everything.
02:12
So we're always going to keep it constrained to the ones that we're looking at individually.
02:16
Another cool thing here is if you were to go and you were to change any of these options.
02:19
So if I was to change the citation style, you can see that there's this little yellow dot showing up that's showing me that I've changed from the default.
02:27
You can now reset to the default or apply this.
02:29
You can continue making changes to the overall settings, going through each of the tabs, changing things as you want to, going in here, changing whatever you need to, and you'll see that these changes are being tracked.
02:41
So this makes it easier for you to know what you changed.
02:43
And if you want to reset or if you want to hit Apply and have these applied settings take place.
02:49
Now, these are saved to the individual session that you currently have.
02:52
So if I refresh this page, I will still have these changes made.
02:56
But we're also looking at the next step here of continuing to improve, allowing you to name setting configurations.
03:02
So if you have specific models you like, specific dashboards, you're looking at, specific, sources you want to look at, certain journals or publications, maybe you want to save those into name settings.
03:13
So we're going to continue to improve this, but we're really excited about this because hopefully it helps make it a lot easier for you to find what you're looking for.
03:20
And definitely recommend always taking time to modify the settings to get the right output that you want.