NEW: GPT-4o Mini + Github Copilot FREE Alternative👨‍💻🤖 Continue.Dev With Qwen2 (VS Code w/ AI)

E [Music] what's going on everybody Josh pook here and in today's video we're going to be taking a look at a few different things GPT 40 mini just was released recently by open AI and it is their cheapest uh most efficient model for you know maintaining efficiency with cost Effectiveness and you can see the evales here we're going to dive into it take a look at that show you how you can actually use it and we're also going to be looking

At continue Dev um an extension in vs code it can also be used in jet brains as well and how to actually set this up so you can start using this as an alternative to tools like uh GitHub co-pilot or even AER or uh Claude engineer I did a video on AER yesterday we'll be doing more videos on Claude engineer Etc so show you how to set this up you can leverage this with any model of your choice gpg 40 mini

Claude Sonet 3.5 even local models we're going to show you how to use it with oama and any model you choose so let's Dive Right In okay so if you're not familiar with continue Dev I'll leave a or I guess it's continue continue dodev though um I'll leave the link in the description to their website you can check out their docs you can check out their Discord and I'll leave a link to the chat GPT 40 mini um blog Post Release

That they did essentially what you're going to have to do where you're going to have to um to actually use it you're going to have to go to platform. openen a.com and you're going to have to get an API key so you're going to go to your dashboard and log in and you're going to get an API key uh you just go to the API key section get that API key put in a notepad secure it because we're going to use

It in just a moment all right now what you're going to do is we're going to go ahead and open a VSS code new window here okay let me just open this and we'll just go new uh open actually no oops we're going to open a folder I'm just going to create a random test folder for the case of this video all this test okay and going to open this select okay so now we're within our test project what you're going

To do is I mean you can use this with any um I mean you can use this with jet brains or vs code um we're going to be using VSS code and what you're going to do if you're not familiar with VSS code you go to extensions here and you're going to search continue okay once you do that you're going to see continue right here this is their extension you can simply just install it you can read about it here uh

I believe they have a GitHub or actually I don't know if they have a GitHub I don't know if they yeah I don't think this is they just have the extension here so you can read that here check out their website and they also have docs in like the community like I said so if you have any issues there was a few issues that I've been seeing um with a couple things um I believe they let us know in the Discord

Like there was a few bugs um some of them are being fixed but um this tool works really well from what I've been using it so far for and it's very useful um so yeah once you install that you will see it pop up over here on the left hand side it will be one of these icons I move moved mine over here right so I just simply dragged it over here to the right uh panel but for the case of

This video just cuz my head is right here I'm going to drag it back so we can see it on the left hand side here let me go here continue so once you're in continue Dev it's going to look like this you can see I was chatting with it already but um if I wanted to do a new session I could do that all right so you're going to do is you're going to press on the plus button all right that's

Going to give you the option to select a model so you can start with a provider or select a specific model so you can just go ahead and click for this case open AI but you can see here we got a lot of different options Gro together AI uh olama local uh local um hosts or local llms Etc so a lot of different options and it even looks like they have their own um something here I oh that's a proxy uh

For other services it looks like yeah so anyways once you do that you'll be able to select like gp4 for example or gp40 but you're not going to see GPT 40 mini just yet they're probably going to update it very shortly but you're going to actually have to go to config to open config right here and and you or you could either just scroll down click this or just click on this setting here and then configure but excuse me but what

You're going to do is you are going to see right here um you'll see model and most likely for you it will say GPT 40 you're going to add a dash and then add a mini you can do the same thing with the title here you can see the context length and you can see the system message right here so you could change this if you like here is where you're going to input your API key I just blanked mine out

Put some dummy text right here just for the sake of this video you're going to see the provider open AI if you're using a different provider you can change this um but you could also just change it here and then enter the model names okay um you can see custom commands so this is I believe just dummy text right here pretty much or cohesive set of unit tests for the selected code okay so now it actually does have this is yeah

And but you can add more or change this and then tab autocomplete model so right now I'm using to by oama or using through oama um not by oama um so I'll tell you how to get that set up if you want to use that you don't have to use that you could you could use gp40 mini but the reason why I'm using a local model is because just for autocomplete it makes more sense just because you're going to save some

Uh it's going to be more cost efficient and you can leverage these open models that are actually really good especially some of the smaller parameter models because um you really any machine can run them almost um if you want to know what the what like the what uh requirements you should have for your system to run um the a specific model maybe if it's bigger you can just ask chat gbt or perplexity but you should be able to run like something

Like this the 1.5b uh model so I'll show you how to do that um and also too if you were to run um like GPT 40 mini you would just add another API key value uh and um spot right here and then you would just put your API key there okay now I'm going to show you how to um get quen to olama as well as I'll show you how to also uh if you do want to use um you know

AI within the command prompt when you're coding I'll show you how to do that too in just a second okay so with uh continue it's really cool because you can ask it any question right here you can ask it follow-ups but you can also add content so you click here and you could select um files right any file you have in your codebase you could select code actually this is files like in your um thing but uh files code docs all

Right get diff um so you can reference the current get diff uh terminal reference contents to the terminal problems reference problems problems in the current file folder uh or automatically find relevant files in your entire code base all right so really cool you could also add more context providers so I really like how they have that option um as these tools these code AI developers get better and better I know that like probably within a year from now maybe even sooner

It will just be like it can read the whole thing or the context or they're going to have some crazy way so um to for these AI models to have more context without maybe leveraging like a hybrid of who knows like rag mix with like larger contexts m with like them releasing more powerful models and finding different whatever uh ways things work efficiently but who knows but I I definitely even right now it's super super helpful for developing and coding but

Anyways so now you know how to use continue let's say you wanted to do some AI prompts within the command line just as a side note uh you could use this tool called GPT shell or shell GPT and you literally just run this installation uh pip install shell GPT so if I was over in my test I go new terminal right here and then I could just copy this now here and you will see that it got installed then once you

Have it installed you can actually go to your uh main directory of your user whatever you're in and then conf doc config shell GPT and then the file is sgpc uh or you could just use the command line to open a notepad with it or a Vim or a whatever um but yeah you're going to open this and from that point you can add your open AI key right here and you can also put the default model right here so for

Me right here you can see I'm using gp40 mini all right but you could use whatever model you like all right and then from that point on I'm not going to save this we go here we can do like this so GP uh sgpt what's up or how's your day going and and you can see here I'm ready to assist you with your questions right here how can I help you today all right so there we go it is running all

Right now me personally I prefer continue um to this shell gbt uh just I mean I just like using this it's a little bit nicer a little bit sleeker but this is cool as well all right and then also lastly um showing you how to use some something like AMA um you can use any model all right for for me I was using quen so quen 2 actually so you would go to ama.com I'll leave the link in the description you

Download olama if you don't have it already right from that point on you go to models up here and then you can select any model you want you can select Gemma 2 uh llama 3 deep seat coder I'm going to be doing more videos on all these different tools quen 2 so for us we'll select quen 2 and then you can select what model you want um or what size you want as you can see I selected this one it's only

Under it's under a gigabyte but you know these bigger ones are about 4 gabes so we're using this one right here and it's pretty good so this is uh you can read about it too you can see the performance compared to llama mol other tools all right so once you actually have the desired model that you want let's just say we're using this one you would just literally copy this Command right here for whatever model you're using and you would just

Run it in the command line right here it would download the model and from that point on then you could go back to continue extension and then put it in the config file like I showed you you need to re go back and see what I actually put there you can just rewind the video again and check take a look there but you literally just do exact ly that and then you'll be able to use that model for autocomplete and then

You can use gpg 40 mini for um generating or asking out questions about the codebase ETC so I've been finding it very useful I've been using it uh a lot recently and using it for different projects and it's very very helpful so I hope you got some value from this video guys if you're new here make sure to check out some of the other videos and if you got some value then make sure to subscribe and stay tuned for future videos

We upload videos every single day on AI uh open source you know AI tools automation marketing business coding Etc um all this to help you grow your business grow your knowledge around AI so I got a lot of cool stuff in the works and plan so stay tuned for that also too guys if you are new here and you haven't already joined our free community strid community.com link for that will be in the description Down Below guys we have about 750

To 800 members in our free Facebook group free Discord channel so you can check that out and then lastly too guys if you haven't already checked out my book client acquisition code you can go pick up your copy link for that will be in the description Down Below guys all the links I mentioned will be down there too so other than that guys I will see you in tomorrow's video keep hustling keep grinding and of course guys accelerate your stride take

Care