Up next in 10
LocalSite: NEW Deepseek Coder! FULLY FREE AI Coder is INSANE! (Opensource + Local)
Sep 16, 2025
https://github.com/weise25/LocalSite-ai
Check the full description to see all Payment Links to Buy Source Code:
Buy the Source Code with PayPal (in USD):
https://www.paypal.com/ncp/payment/5AGGYJECQVZJC
Live Demo of App:
Buy the Source Code with Stripe (in INR):
https://buy.stripe.com/5kQ4gB8rCbcydjD4Vm14f1J
More Payment Options:
Watch My Visual Studio Code IDE Setup Video For Fonts,Themes & Extensions
https://youtu.be/Bon8Pm1gbX8
Show More Show Less View Video Transcript
0:00
live stream. So in this live stream,
0:01
I'll show you a new Deepseek open-source
0:05
coder on the command line uh which the
0:09
Deep Seek company has developed. It's
0:10
Deepseek. If you don't know, it's a
0:12
Chinese-based open-source model.
0:15
Specifically, it's very popular for
0:16
coding. So, this is their official
0:18
website. If you simply type DeepSync,
0:21
so deepse.com right here, you can check
0:24
out their chat prompt. So they offering
0:26
this uh open-source uh command line
0:29
interface where you can generate web
0:31
pages components with s simple text
0:33
prompts. You can also integrate it with
0:36
uh any local model that you have
0:38
installed or downloaded or you can also
0:41
integrate cloud models as well. So this
0:44
is their GitHub repo right here and uh
0:47
it's basically developed inside NexJS,
0:50
React, Tailwind CSS.
0:53
So now to basically get started, you
0:56
clone this repo. First of all, simply
0:59
open your command line and
1:04
right here you simply clone this repo.
1:09
Just paste this command.
1:14
So this will install the repo right
1:16
here. Then just go to in the repo local
1:19
site AI and then simply run this command
1:22
npmi. So this will install all the
1:25
necessary packages uh which are
1:27
required. So
1:30
so once it get installed right here you
1:32
can open your env.local file here you
1:36
can replace your API keys.
1:38
You can go deepseek API key or open AI
1:42
API key as well. So
1:45
if you're using any sort of local
1:47
provider, you can provide this provider
1:50
name right here.
1:53
So if you want to use the local, then
1:57
you simply paste the URL where your
2:00
local model is running. So
2:05
just wait here. Let me just configure
2:07
this and show to you how to set up this.
2:10
So now once it is installed so we can
2:12
simply open it inside VS code. So this
2:14
is the overall nextjs project right
2:18
here. So it's uh now it's very simple to
2:20
run this. So
2:23
we simply run the command here npm
2:26
rundev. So this will start your
2:28
application at localhost 3000
2:32
and uh then now just configure this API
2:36
keys. Simply open this file uh env
2:42
uh
2:44
for this you just need to create this
2:46
file. First of all,
3:10
So right here you will see this env file
3:14
here env. So right here you just need to
3:18
uh
3:20
replace the API key. So if you're
3:22
working with deepseek right here you can
3:24
replace your API key. So once it starts
3:27
your application, you can open this at
3:30
localhost
3:31
3000.
3:35
So this will actually open your
3:36
application. It is simply compiling it.
3:38
So just wait. So I've given the link of
3:41
this repo in the description of this
3:43
live stream. And uh
3:47
so now it will simply load this deep
3:50
coder. So right here inside the chat
3:53
prompt you will simply type what sort of
3:55
application you are generating. So right
3:57
here you can select the provider. If
3:59
you're using Olama you simply select
4:02
Olama right here and simply select which
4:04
model you are using. It will
4:05
automatically detect if you have already
4:07
installed Ola inside your system. So
4:09
I've already installed this model
4:11
through. So it's automatically detecting
4:13
it. And right here we can simply say
4:17
please build
4:19
a image to PDF web app in browser. So
4:23
you simply write the prompt and click on
4:25
generate and this will be fully fledged
4:28
coding editor like VS code. So the code
4:32
will be generated and you will see the
4:34
live preview uh on the right hand side
4:37
in various screen devices. You can say
4:39
mobile, desktop,
4:41
tablet.
4:45
So it's actually mod uh initializing
4:47
your model.
4:50
So now it's successfully writing the
4:51
code. Here you will see that.
4:56
So basically I'm running the local local
4:58
model through Olama. So for this you
5:01
should have at least one model
5:02
installed. If you just want to start
5:04
completely free go to this website
5:06
ola.com and just search for any
5:09
open-source model. For example, I
5:11
installed this model here deepseek R1
5:14
which is completely free to install and
5:17
I downloaded this by running this
5:19
command. So you should have installed
5:20
this.
5:22
So I have already installed any one of
5:25
these models you can install are
5:27
depending upon the size. So that can run
5:30
in any machine. So you can see it
5:32
actually generated this uh you basically
5:35
upload any sort of image and uh
5:40
it also show you the live preview as
5:42
well. So
5:46
you can even ask modify or ask follow-up
5:49
questions as well. So
5:54
you can also refresh
6:01
So here you select whatever image you
6:04
want to convert.
6:06
Click on convert to PDF and your image
6:09
will be converted to PDF.
6:12
And there is also this option right here
6:14
uh which allows you to export your code.
6:16
Click on export and then it will give
6:18
you this HTML file which is generated by
6:21
this AI provider. So it generated all
6:24
this code here and it also generated
6:27
short little summary as well.
6:43
So here you can even edit the code as
6:45
well.
6:47
So
6:51
you can build out any sort of
6:53
application. So right here I will say
6:56
build a scientific
7:00
calculator with premium UI in browser.
7:06
So as you can see it's using this Olama
7:09
uh opensource
7:11
local model that I have installed here
7:13
and uh can see again it will start
7:16
generating my scientific calculator. So
7:20
it's a fully fledged you can run this
7:22
locally so there is no need to actually
7:25
paste any API key or any sort of token
7:28
limit. So this is completely free
7:29
unlimited access. So once you clone this
7:31
repo uh I have shown you the
7:33
step-by-step instructions. So you should
7:35
have at least one local model installed
7:37
through Olama. Simply install this
7:40
automatically that will detect it and
7:42
then you can start building out any sort
7:44
of coding applications through single
7:46
prompts.
7:47
So it's a really insane open-source deep
7:50
CKI mod uh coder.
7:56
So it's basically having this interface
8:06
So this is very basic model that I've
8:08
installed guys. That's why the quality
8:09
of the results that you are seeing right
8:11
here is not professional. So that's why
8:14
if you want to run uh the biggest models
8:16
such as this one which is more powerful.
8:19
So the sizes you can see. So majority of
8:22
the computers can't run this models
8:25
because uh the computer needs to be
8:27
really uh intensive then only it can run
8:31
these big models. So so I'm just showing
8:35
you how to run this. If you have a
8:37
powerful machine then you can install
8:39
the biggest model possible to actually
8:41
get better results. So depending upon
8:44
the model you may get uh ordinary
8:47
results something like this. But this is
8:49
the application definitely try this.
8:52
I've given the link in the description
8:53
of the video.
#Programming
#Web Design & Development