WEBVTT

00:00:00.380 --> 00:00:03.969
Now in this chat, no matter
if it's here in the sidebar or

00:00:04.000 --> 00:00:07.460
separate window,
you get various things you can change.

00:00:07.480 --> 00:00:11.320
And again,
the exact UI will also change over time

00:00:11.380 --> 00:00:13.120
features, of course, are the same.

00:00:13.140 --> 00:00:16.840
Uh, for example,
you can pick the AI model you wanna use

00:00:16.940 --> 00:00:20.900
just as before,
you can also add new models here by

00:00:20.940 --> 00:00:24.920
clicking here and be taken to the, uh,
options here where you

00:00:24.960 --> 00:00:28.880
can register new models. And as new models

00:00:28.960 --> 00:00:32.580
become available,
some models need to be explicitly enabled

00:00:32.619 --> 00:00:36.420
available in your chat.
So you might wanna check that

00:00:36.540 --> 00:00:40.449
which you would like to use.
Many models also have this

00:00:40.600 --> 00:00:44.380
deep think mode here, this brain mode,
which means they

00:00:44.520 --> 00:00:48.240
can maybe tackle more complex tasks.

00:00:48.260 --> 00:00:52.240
And then there is this max mode
which means you're fine with including

00:00:52.300 --> 00:00:56.160
more context, more files,
more information, which can

00:00:56.200 --> 00:00:59.480
produce better results but
which will also burn more

00:00:59.540 --> 00:01:03.320
tokens, which are, in the end,
the fuel for which you're paying every

00:01:03.400 --> 00:01:07.060
month. So depending on your plan,
you have more or less Cursor

00:01:07.180 --> 00:01:10.880
usage included. And with max mode,
you'll burn through your

00:01:10.960 --> 00:01:14.840
allowed usage quicker,
which means you might need to, uh,

00:01:14.880 --> 00:01:18.740
top up, you need to buy more tokens,
or wait for the next month to

00:01:18.760 --> 00:01:21.060
begin.
So you'll wanna be careful with that.

00:01:21.140 --> 00:01:24.960
Auto means that you let Cursor choose
which model to use

00:01:25.060 --> 00:01:27.940
for a certain prompt, and whilst that,
of course,

00:01:28.340 --> 00:01:31.920
removes a lot of the control you have,
this

00:01:32.000 --> 00:01:35.650
actually is a mode
that gives you unlimited usage

00:01:35.660 --> 00:01:39.520
Because for the top models, there
is limited usage, again,

00:01:39.540 --> 00:01:43.160
depending on your plan. In auto mode,
Cursor will use cheaper

00:01:43.170 --> 00:01:46.920
models behind the scenes whenever it makes
sense, and therefore they guarantee

00:01:46.960 --> 00:01:50.490
you unlimited usage because they can
always fall back to these cheaper

00:01:50.520 --> 00:01:54.410
models.
I typically use max mode with the top

00:01:54.440 --> 00:01:57.690
have the premium plan, and
that gives me the best results.

00:01:57.700 --> 00:02:00.520
But you can tweak this
and adjust this for your exact

00:02:00.560 --> 00:02:02.060
requirements.