Up next in 10
Run ANY AI Model on Your Machine WITHOUT a GPU! (Ollama Cloud)
Show More Show Less View Video Transcript
0:01
Uh hello guys, in this live stream I
0:04
will actually show you uh a technology
0:06
by which you can actually run any AI
0:08
model directly inside your machine
0:10
without having to actually requiring a
0:13
GPU or any kind of hardware. So this
0:16
software is called as Olama. Olama is
0:19
really popular.
0:21
If you simply type right here, the very
0:23
first link which appears uh this
0:25
actually is a desktop software which you
0:28
can actually directly install in inside
0:31
your machine as a desktop application.
0:34
So it also allows you to download any AI
0:37
model directly inside your machine as a
0:39
local model. You can directly run the
0:41
model locally. If you go to the website
0:43
ola.com, uh if you go to models, it has
0:46
all the models available to you. You can
0:49
search for any favorite model that is
0:51
you want to download locally. So you can
0:54
see the name of the model.
0:56
You can directly install this model
0:58
locally. But there is also a second
1:00
option that you can also run it using
1:03
their clouds. So, Olama cloud is
1:06
actually if you don't want to download
1:08
this model which is 24 GB and uh running
1:13
it using your own GPU, you can actually
1:16
uh if you simply
1:19
type Olama cloud
1:22
this actually allows you to
1:26
so all the models are also available
1:28
inside their clouds. So rather than
1:31
downloading the models, you can actually
1:33
use the cloud version very easily. So I
1:36
have already downloaded Olama as you can
1:38
see it's actually a desktop software. So
1:41
once you launch Olama right here inside
1:45
this section in the chatting section you
1:47
see this cloud section right here. Uh
1:49
rather than downloading that model you
1:52
can actually use the cloud version. So
1:54
simply select your model whichever model
1:56
you that you want to use right here.
1:58
Let's suppose I want to use Kimmy K2
2:01
2.5. After selecting that model, simply
2:04
send out your message.
2:08
So whatever message that you want to. So
2:10
let's suppose I want to build out an
2:12
application.
2:14
So I simply type the prompt here. So
2:17
after selecting this, you will see the
2:19
model will be really fast because rather
2:21
than downloading this, it's actually
2:23
using its cloud version. You can get
2:26
started for completely free here. As you
2:28
can see, it's using the Kimmy K2.5 model
2:31
right here. Uh it's generating the code
2:34
that you have uh asked it to which is
2:37
image to PDF.
2:40
So it's really fast. It has all these
2:42
cloud models which are there. You don't
2:45
need to do anything. Just select the
2:47
model and instantly you will see the
2:50
result here. It will actually generate
2:52
fullstack coding applications. It can
2:54
actually research the internet. All
2:57
these things it can do.
3:01
So this is a great way by which you can
3:03
actually run any AI model without if you
3:06
don't have a powerful machine or GPU,
3:08
you can still run any AI model inside
3:11
Olama using Olama cloud. So first of all
3:14
download software. So this simply
3:18
install this and as you see in the
3:21
pricing in the free plan they do allow
3:24
you to use the access cloud models their
3:28
CLI API desktop app.
3:32
So
3:34
if you want to see the how much access
3:38
you're remaining
3:41
so you can see it's still generating all
3:45
the stuff here. So this is actually the
3:47
functionality cloud models in Olama with
3:50
the help of which you can run any AI
3:52
model for completely free.
3:55
It also have the download feature as
3:57
well. You can download the model or run
3:59
it through their Olama cloud.
4:02
So you can see it's actually generated a
4:05
full stack image to PDF coding
4:07
application is inside 5.7 seconds. So in
4:11
this easy way you can see that
4:18
You can simply say top frameworks for
4:22
you can ask any question right here.
4:24
Again you ask the question again it will
4:26
list out all the frameworks
4:29
and you can just see here how the speed
4:32
is really fast here because it's
4:34
actually using the cloud models rather
4:36
than downloading the so it really will
4:39
be First
4:57
and if you want to see their usage, you
4:59
simply go to the settings and just see
5:02
how much usage you allow. So it allows
5:05
up to free usage. So right here you will
5:09
see whenever it's completed 100% you
5:12
will get a notification but the free
5:14
version allows you up to very good uh up
5:19
to 100,000 messages are allowed in the
5:23
free plan. So it's more than enough. So
5:27
if you do want to upgrade you can even
5:29
upgrade as well. Their paid plan start
5:31
with $20 per month but the free plan is
5:34
more than enough. You can simply select
5:36
any cloud model from here and simply
5:39
send out the message and it will be
5:41
really fast because it's actually using
5:43
the cloud model right here. So
5:47
definitely check this out and start
5:49
using this.