I created this amazing high-quality production-ready website with all of these pages and e-commerce functionalities within less than one hour using this amazing AI code editor which is called Bolt. New. And I'm actually really amazed by the performance of this uh AI code agent, but

the problem is I did create that website within less than one hour, but the problem that I had was actually I consumed so many tokens. So Bolt. New actually gives you 10 million tokens uh for less than $20 uh per month, but I consumed 5.1 million tokens on this one

project only. That means if I have to create more projects in Bolt. New that I really really want to, but I it will actually cost me a lot of tokens to like um build even like something like a landing page or maybe an app or something. So I really really needed a

solution. I love Bolt. New, but it's too expensive for me. I might end up like maybe $200 a month because I just love this tool so much. So I was like okay, I need a solution. So I looked for like how we can run this Bolt. New locally on our

computer, just like any large language model. So I was like okay, uh maybe because they're actually an open source, so I looked at their uh GitHub uh repository. So they have like their open source, their entire source code is available online on GitHub. And then actually I searched on YouTube, I found

this video. This guy, he did an amazing job. He spoke about Bolt. New and he created a new version, and this is Bolt. New, any LLM. So what this actually does, I'll actually leave a uh link to this uh repository uh in the comment section so

please uh just you do not discontinue here because actually I'm going to give you a very detailed step-by-step guide of how you can actually run this locally on your uh computer and then use any large language model like free ones like uh like you can use Llama 3.2 you can use uh

Quynh 2.5 coder or any other large language model of any size of any variant you can use that and then uh like write code using Bolt because the problem with Bolt right now on the on on B. new it doesn't give you the option to

change uh large language model here so you you only get access to like uh Claude sonit 3.5 I think and you cannot get anything uh like any other you can if you let's suppose if you want to like code with uh GPT 4.0 so you cannot do

that um with uh B. new input production item so that's why we need actually this to like uh deploy this locally on on our computer and then use any large language model that we want to run um to like make more efficient code so before we

start with that please please do not forget to subscribe to my YouTube channel it is really really I it takes me a lot of time to like make these kind of videos and then do all of this research and then share my experiences with you so please if you think that you learned

something from me or if you think in general that my tutorials are good so please just do not forget to subscribe to my YouTube channel so what you have to do is uh you need to uh if you're new to programming or if you're new to like a

large language model or AI development so you don't know what is GitHub, you don't know what is a GitHub repository. So I'm actually going to walk you through in a very detailed step-by-step process. So the first step is, uh, you need to come here onto this repository. I'll

leave a link uh in the description so you you'll just have to come onto this GitHub repo which is Bolt-New-NM and then you can do click on code and then click copy your uh this GitHub repo link. Next step is now you have to

install this repo locally on your computer. Either you can just directly like download as a zip or the best uh option is as a programmer or if you're not a programmer, you want to be a programmer or you want to like write programs uh like code using AI so I

recommend you like then act like a programmer. So what you have to do is just uh open your CMD command prompt and then move to a directory like uh I don't know like just any create any folder anywhere on your computer and then just

move to that directory. So I have actually already installed that on my computer. Let me just open that. I think it was in my development section and in here Bolt-New-Local-LLM all right so this is my directory so what you have to do is just

copy this directory link, go to your CMD and then what you can do you can do this to CD and then enter this if it doesn't uh goes to that directory, what you have to do is f and then write these colon and then just click on that so as

as soon as you're here as soon as you're inside your directory then you can just write get clone c l o n e and then what you can do is actually you can just copy um the this thing and paste it here on there and then just simply click

enter and it will actually start uh cloning this GitHub repository as soon as it will actually clone it will just create uh this folder and every all of these files inside this folder in here uh right inside your folder so you can just like come here so now next step is

uh just use any uh IDE I highly highly recommend using VS Code so VS Code is actually a revolutionary uh uh like uh AI code editor I highly highly recommend you using that if you are new to programming and writing code and if you don't do not want to like waste too much

time in in um learning everything from scratch and I I highly highly recommend you using VS Code so what you can do next is just open this directory uh come to VS Code and then click on file open folder and then go to the your your directory where you clone your

repository and then simply just click on that and just simply select folder just click on that and it will actually open something uh this is what you'll see like initially so this is this is how it will open so you can just simply open your VS Code setup VS Code everything for

you and then our next step is actually in your terminal what you have to do you have to write something uh npm hold on just a second hold on just write npm install and then click on the enter as soon as you write that it will actually install all of the

dependencies and it will actually create this node_modules folder for you and it will install all of the dependencies, all of the packages that uh this uh repository is actually using. Once you have done that, uh it will take maybe like 3 to 4 minutes to install all of

these dependencies. It depends on your internet connection. The next one is actually, you can just npm run dev. Just type that and then click on enter. Once you click on enter, what it will do, it will actually run um your project locally on

your uh desktop. So just simply click on that and now it will show you this link. Just simply open this link so you can see that now our B.new is running locally on our computer. Next step is now to use uh a large language model and then write

um like use that large any large language model for writing code. So you can see inside here you'll see find this uh .env file. Let's just rename it to env.local. So simply rename it to that. And the next step is before we even like do

that uh okay, so all right, so uh here inside this uh env file you'll find all of these like um options to add API keys. So let's suppose if you want to use OpenAI, so just go to OpenAI API key and uh get an API key from here.

But this will definitely cost you, so like this is not going to be free. So I mean, still it's it's a quick and easy option to just like get started with uh using uh uh using B.new locally on your computer, but still it's it's going to be

expensive for you, uh, open API key so you can just like simply have it from here. Or, the next step, all I think you can just, just go to dashboard and then create a new API key. The next step, uh, the next thing that you can do, like you can even use

entropy CLA or um any of these um um like uh large, any of these large language models. You can even use MRoL or any of these large language models you can use here. But what we're interested in, we want to use uh free one, uh free

any large language model that we can run locally and then we can just like use that large language model to uh write code for us using Bolt.new. So like imagine like having access uh to Bolt.new without paying any single uh penny for that. So what what you want to

do, just go to something like AMA. Just go to Ama or maybe LM Studio. So if you're aware of AMA, so this is just like a kind of large language model management uh uh tool. So consider this like just your CMD for large language models. So just go to model and then just

copy uh and let's suppose yeah, Qu 2.5 coder, this is a brilliant, brilliant large language model right now for writing code. I have been loving it. So you can use any of this variants, you can use 7B, 14B, or 32B, depends on what your

computer actually supports. So if your computer supports, I mean if if your computer has enough uh uh hardware requirements, if if it has a good uh kind of GPU, so you can um use uh maybe 32B and install it directly on your computer. It won't slow, slow on your computer but

if you have a low-end computer it will actually definitely slow down on your computer so I recommend using maybe 7B I have installed that already on my computer so as soon as you click on that just copy this thing uh just like this command and then come to

Powershell make sure you are running AMA already so AMA once you install AMA it will actually install like this so if you have not installed it in your computer first first install AMA download click on the download button it will download this uh AMA download x file double click on

that and then install that and then it will actually uh install in your computer then next you can just open AMA and then you can see in here it will be running on your computer it's not running right now so let's just click on that and it will now start running in

your computer you can see in here it is running so next step is actually you can just click on Powershell so just go there uh let me just okay so in Powershell all you have to do is this command that you copied from here just paste this command here and then

click on enter as soon as you click on enter it will actually take couple of minutes and then download and install uh this coin 2.5 coder uh directly in your computer locally once you have done that next step that you have to do which

is very very crucial step that I think is mentioned on this GitHub repository as well which is uh let me just scroll down scroll down and this one so what you have to do you have to create a model file folder inside your uh uh your inside your inside your

project so right here inside your project I have created a folder which is called uh model files which is that where is that yeah here it is sorry uh I'm sorry so uh model files just just uh create a new folder named model files and next up is just create a new file

named uh Quen 2.5 coder uh so if you're using Quen 2.5 coder use this name if just like using Llama 3.2 Vision or any other variant of Llama 3.2 or any other model or maybe something so just use that

name here what next step you have to do is just copy this thing paste it here and now from you have you see this from and then the next thing is Quen 2.5 coder uh 14b so just uh write your exact um

like variant from here so if you're using 14b use 14b if you're using 7B just use 7B I think I'm using 14b so I'll just use 14b so just copy it from here uh this part only just copy it from

here paste it here and then that's it uh the next next you can just leave the next line as it is next what you have to do is you have to now uh navigate to this file just like uh CD uh if you're inside your uh project so let me just

uh show it to you how I did that yeah so just go to your uh like wherever your project is located and then inside that just go to your CD bold. new this was my project directory and inside that there was this folder so move to like make sure you're inside

that folder and then run this command which is AMA create then the F and then a Quinn 2.5 coder and then name your uh this LLM uh this is a custom name that that I gave to this LLM and then 14B so

just uh run this command this entire command you can see this entire command so run this entire command uh in here and it will actually uh create this uh like uh install everything for you uh like assign that yeah that large language model to uh Bolt. new local

version so as soon as you have done that uh now you can actually come here and you can uh like uh as if your Llama is running and you can see uh you can access that large language model right here in your computer uh hold on yeah here it is so right now I have

AMA and then uh where is here is the 14B so you can see uh any uh LLM any large language model that you install on your computer you can run that like you can access that now here and then you can use that to write uh a code without uh

like costing even a single penny to you uh if you if you're going to use maybe okay so another way is actually to other than using AMA you can use LM Studio which makes things lot simpler so you can just go to uh your uh browser uh go

to LM Studio there's LM Studio okay let me just go to there LM studio just go there uh open that download it for Windows and then install it and then open it on your computer and if your LM studio is open and the next step is you can just go to this

search directory and then download any of these large language models from here you can use coder uh, Quin 2.5, coder 14B uh, and then just download that, install that, and the next step is now you can actually access that inside your uh thing here so I haven't I think uh

installed any large language model for LM Studio so far so that's why it's not showing anything here so yeah uh that's how you can actually run um this model directly in your uh on on your computer without uh anything and then

you can actually uh run it uh using any large language model and then this way actually you won't have to pay any of these fees so you can see like you can easily consume 120 million tokens if you are using if you love B. new you can

easily consume 120 million tokens and then you end up paying like $200 just like me so do not do that and then install B. new locally on your computer and then run it locally and then develop this kind of like high and high quality

applications directly in your computer um I'll Next Step, next video I'll make actually an in-depth tutorial for B. new but yeah I mean um if you have really uh if you're still watching this video please just do not forget to like and subscribe to my YouTube channel this is

really important and it gives me a lot of motivation when someone like subscribes to my YouTube channel I'm a small uh YouTuber still so I just need more subscribers please so I highly highly request you to please just subscribe to my YouTube channel and if you

Have any ideas that you want me to, like, make videos on? Just let me know and I'll actually create videos on that certain particular topic. I'll see you in the next video. Bye-bye.

I Forked Bolt.new AI Code Editor and made it way better @ColeMedin

The Metaverse Guy

Bolt.new를 사용하여 1시간 이내에 고품질 웹사이트를 만들었습니다.

510만 토큰을 소비하여 월간 한도를 초과했습니다.

향후 프로젝트의 높은 비용에 대한 우려.

Bolt.new를 로컬에서 실행할 수 있는 솔루션을 찾습니다.

Bolt.new-anyLLM을 소개하는 YouTube 동영상을 찾았습니다.

Bolt.new-anyLLM 리포지토리에 대한 링크를 제공합니다.

Llama 3.2 및 Quynh 2.5 코더와 같은 다양한 LLM을 사용할 수 있습니다.

특정 LLM을 사용해야 하는 Bolt.new의 한계를 극복합니다.

로컬 설치에 대한 단계별 가이드

GitHub에서 Bolt-New-NM 리포지토리에 액세스하세요.

Git을 사용하여 리포지토리를 복제합니다.

CMD를 사용하여 설치하고 디렉토리로 이동합니다.

VS 코드에서 프로젝트를 엽니다.

npm 설치를 사용하여 종속성을 설치합니다.

npm run dev를 사용하여 로컬에서 프로젝트를 실행합니다.

.env 파일의 이름을 .env.local로 바꿉니다.

OpenAI(유료)와 같은 LLM용 API 키를 추가합니다.

AMA 또는 LM Studio에서 제공되는 것과 같은 무료 LLM을 사용하세요.

AMA를 사용하여 Quynh 2.5 코더를 다운로드하여 설치합니다.

'모델 파일' 폴더를 만들고 LLM 구성 파일을 추가합니다.

AMA를 사용하여 LLM을 생성하고 Bolt.new에 연결합니다.

LM Studio를 사용하여 LLM을 관리하고 사용하는 다른 방법.

로컬 LLM 사용 시 비용 절감 효과를 강조합니다.

향후 Bolt.new에 대한 심층 튜토리얼을 홍보합니다.

시청자가 좋아요를 누르고 구독하고 향후 동영상 주제를 제안하도록 유도합니다.

I Forked Bolt.new AI Code Editor and made it way better @ColeMedin

I Forked Bolt.new AI Code Editor and made it way better @ColeMedin

스크립트