Omg perfect time of this video, I was searching for a free way to use bolt to build my Saas which give just a little bit of free acess. And I just landed on this video. Thanks a lot
I would like to know how to install all these packets on bolt locally in open sources I have seen a lot of videos but they do not show the installation step by step, as my English is very bad I don't understand well, if you can give us a tutorial to install it on windows 10 or 11 I thank you.
I often receive errors that cannot process my request. no details were returned Often then when the code is created I get red messages saying that the operation is not secure
im so new to this, i have no coding experience, but I'm trying to learn, do you have a personal course on how to install this and all it's updates, and how to keep on updating it? any crash courses? lol
i have installed it and runs smoothly, apart from a few thing, that after building the app, it just stucks on run command at the end, even though it previews the app in the preview tab correctly, The other thing is i was not able to use this with the github models and the gemini api, can i have a detailed video link you're talking about please
thats your job to understand how LLM's work and not feed it too much at once. Thats not a "bolt" issue, its not understanding how LLM's work. break up your tasks, read the error log, look in your browser console.
What setup would you recommend for a project using next.js, shadcn and supabase? It can be paid or free just looking for the best out there. To create components i'm thinking about v0 og bolt, but that's only to create a starting point. Have watched a lot of your videoes - great work 🙏
I have messed around with this and it isn't so easy. And the bigger problem wasn't the APIs it was that it is Isolated to one docker container. And one local file. Whatever you thought you wanted to do would probably be better executed elsewhere while it is under development for that reason.
Let me briefly explain, if I want to train something to do a task I would have to a. Train it on the location build it's memory and then implement a solution. There goes 1mil tokens. Now it knows kind of how to do it. Run it again and it has to recall the memory hopefully execute the task and it can only do it from one file and one location. There goes another 150k tokens. I think you can see where this is going.
Even then you will be limited to one local file and a docker container unless you are a advanced user who can configure everything to your needs. I am a huge fan of the project. But a very pessimistic believer that it will actually work. Advanced knowledge is a must to do anything because there are no guard rails.
If they built in guard rails for system protection ie. The AI can not delete my system etc. then we would be cooking. But without basic guardrails agent zero can't be exposed to the larger data streams. Which means it can't be trusted which means it is not truly functional. Feel free to disagree but setting loose a AI that can do anything is not what I want to see.
I was about to ask why didn't you cover cole melin's fork. Though I must say the paid version is still far more stable than open source one. I guess bolt. New they got some secret behind the scenes to prevent AI builder from going crazy when dealing with more complex apps.
I noticed that they prompting is straightforward and is far more better in terms of prompt engineering and it makes sense since these models have been sophisticated through times. Bolt.new is best for outputting better result in terms of full stack apps ( I am using the free version by the way)
More improvements are comming to open source. I have second most popular fork and we will be joining forces with more people to start iterating on the fork quicker.
Do you know of any models or open source apps that focus more on creating better UI's like v0 instead of focusing more on functionality?? That would really help me out.
nice work 🎉. just a question please: how api for chat-gpt4 is free via github? can you explain please? or you mean there is a limit rate?! thanks once again 🎉
Hello I have tried it on Windows with the Following tasks: • pip install litellm • set GITHUB_API_KEY=xxx • litellm --model github/gpt-4o but then i get this errors: raise ImportError(f"Missing dependency {e}. Run `pip install 'litellm[proxy]'`") ImportError: Missing dependency No module named 'backoff'. Run `pip install 'litellm[proxy]'`
@@AICodeKing my Link looks like this: Uvicorn running on ←[1m0.0.0.0:4000←[0m (Press CTRL+C to quit) It's different to yours. When i start and take "OpenAILike" there is no Model... I hope you know the Answer? This is the env-File: # Get your OpenAI Like API Key OPENAI_LIKE_API_KEY=0.0.0.0:4000
By not using Anthropic api key... instead using Mistral etc...isnt that means poor coding output because Claude is the best coding ai assistant superior to Mistral?
Yes,you will get a much better output if you use it with claude 3.5 or gpt-4o,to be honest it is worth the price that you pay for those api’s especially if you need things fast and professionally looking
@@xavierf2229as I watched on the above video, mistral worked better that gpt4o relatively with the free api keys. Do you have any suggestions what would be the better option for a free api key for more professional coding output?
After setting everything up and running it I get "Mistral API Key:Not set (will still work if set in .env file)" does anyone else have this issue, or know how to fix it? Thanks
@@yusufisroil Beside where it says "Not set (will still work if set in .env file)" there is a tiny pencil icon, click that and put your API key in there (and make sure your on the Chrome browser as it didn't work for me FireFox) then it should work.
Hi, First time here! lost me on min 2:22 .. everything went fine on my OS ,, now ( mv .env.example .env.local ) am stuck lol ,, i can see the file ( .env.local) any help !! again i have 100% zero skills on coding ... lool ,, thanks everyone 💚
Hi AICodeKing, Thanks for another great video. You mentioned a video where you explain how to setup the Github free Models in depth (ua-cam.com/video/p_tyWtQZx48/v-deo.htmlsi=LhgIhTcQUIzFAL3O&t=412) What is the title or link to that video?
Honestly the King is the correct title for ya. Always up to date, always showing the coolest Ai projects. Love ya
You are not just a King, you are a LEGEND man!
Bro you is the bomb!🎉🎉🎉
I celebrate you...
Man Your Video are Straight Forward , no extra nonsense talking , informative 100%
Omg perfect time of this video, I was searching for a free way to use bolt to build my Saas which give just a little bit of free acess. And I just landed on this video. Thanks a lot
Thanks
Thanks a lot for the support!
Bro you really giving us good update i watch your video the mintue it dropped
De todos los asistentes de coding Bolt ganó el único espacio en mi corazón, ahora traes esto! Gracias y Dios te bendiga🙏
'pnpm run dev' for those on Windows (as many AI users are with their GPUs)
thank you!
Mistral Api work perfect with this and also google api
What Gemini are you using? Mine always throws an error with RESOURCE_EXHAUSTED
Hey, what gemini model did you use? Mine always throws an RESOURCE_EXHAUSTED error
@ 1.5 flash or wait few minutes after every prompt for the pro
you have thought me so much within the period of one month ... Great fan here
Dude you are my hero. Thanks from Burkina Faso
Mine doesn't work "Mistral API Key:
Not set (will still work if set in .env file)"
please help
I would like to know how to install all these packets on bolt locally in open sources I have seen a lot of videos but they do not show the installation step by step, as my English is very bad I don't understand well, if you can give us a tutorial to install it on windows 10 or 11 I thank you.
No attachment file 😢
Hello friend, I receive many errors on my screen that say 'The operation is insecure' and the preview does not load. Help me please?
nice vids, with that, possible to code in flutter like in idx but locally that's correct ?
I often receive errors that cannot process my request. no details were returned
Often then when the code is created I get red messages saying that the operation is not secure
ı have same error, dıd you fıx ıt ?
@@furkanyldz1773 No. I was able for one day to use grok but today it gives me error on all.
@@furkanyldz1773 Unfortunately, no
im so new to this, i have no coding experience, but I'm trying to learn, do you have a personal course on how to install this and all it's updates, and how to keep on updating it? any crash courses? lol
i have installed it and runs smoothly, apart from a few thing, that after building the app, it just stucks on run command at the end, even though it previews the app in the preview tab correctly, The other thing is i was not able to use this with the github models and the gemini api, can i have a detailed video link you're talking about please
Here you go : ua-cam.com/video/Y4UuCmZFJZo/v-deo.html
Please how to install it and preview it on windows
its literally in the repo.
This is exactly what I was looking for, I wanna use AI as a quick intern to do dogs body work that bores me :D great stuff!
The AI King with MRBEAST prowess and endurance to be super updated with everything
I'm actually a big fan of the bolt use any llm and the community
cant install litellm using pip or pipx do you know any alternative for that? hate pip, rather use brew or just an app that I can nstall
Does this have better context with not forgetting things like the scope of your whole project or firebase permission errors?
thats your job to understand how LLM's work and not feed it too much at once. Thats not a "bolt" issue, its not understanding how LLM's work. break up your tasks, read the error log, look in your browser console.
@EditMyPodcast Did all that. 500 times.
I setup everything, but when i try to run it it says "Operation is insecure"
hey buddy, is it possible to use bolt with vertex ai api keys? thanks in advance
Contributor here..thanks from my side
is possible to use on windows? and how?
Nice explanation bro!
how to create api from github?
Thank you so much for this!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
8:13 the dancing dinosaurs
Hello thank you for you video I have one question this model we useing localy how it is with private data ? Bolt see our files or maybe code ?
What setup would you recommend for a project using next.js, shadcn and supabase?
It can be paid or free just looking for the best out there.
To create components i'm thinking about v0 og bolt, but that's only to create a starting point.
Have watched a lot of your videoes - great work 🙏
V0 is the answer. All those tools mentioned have partnered with vercel, you can look for V0 demos/usage from official channels or from other people's
Great it works but slow for me, is it because if have only 1VRAM of GPU?
Please answer not just heart it lol
yes, more vram = faster
How to find the free api models page
bro you are amazing, but can you please make a video on how to use agent zero with github's free api and mistral codestral?
I have messed around with this and it isn't so easy. And the bigger problem wasn't the APIs it was that it is Isolated to one docker container. And one local file. Whatever you thought you wanted to do would probably be better executed elsewhere while it is under development for that reason.
Let me briefly explain, if I want to train something to do a task I would have to a. Train it on the location build it's memory and then implement a solution. There goes 1mil tokens. Now it knows kind of how to do it. Run it again and it has to recall the memory hopefully execute the task and it can only do it from one file and one location. There goes another 150k tokens. I think you can see where this is going.
Supposedly 4o mini works better with this framework and I can see why. You need a super cheap or free advanced AI to make it work
Even then you will be limited to one local file and a docker container unless you are a advanced user who can configure everything to your needs. I am a huge fan of the project. But a very pessimistic believer that it will actually work. Advanced knowledge is a must to do anything because there are no guard rails.
If they built in guard rails for system protection ie. The AI can not delete my system etc. then we would be cooking. But without basic guardrails agent zero can't be exposed to the larger data streams. Which means it can't be trusted which means it is not truly functional. Feel free to disagree but setting loose a AI that can do anything is not what I want to see.
so this has a daily rate limit ?
Alright. So what are the Framework this tool supported? I need both Next JS and Flutter.
You can check it out on the Documentation
Is this in linux?
can we use this in Windows?
Love u'r content. Could you kindly list all the available models which has FREE API please.
i have no preview available... any idea ?
Can you do a video on how to use Project IDX for .NET(C#), with a simple notepad CRUD project???
hey man can you create a video on how we can intergrate the mistral API in cursor?
i cant install pnpm
This is an awesome video but is there no option for a Claude api for free with no rate limits
Is the api usage free or does it cost?
And where can I watch the video in question in the api?
Did you find out the way?
I was about to ask why didn't you cover cole melin's fork. Though I must say the paid version is still far more stable than open source one. I guess bolt. New they got some secret behind the scenes to prevent AI builder from going crazy when dealing with more complex apps.
I noticed that they prompting is straightforward and is far more better in terms of prompt engineering and it makes sense since these models have been sophisticated through times. Bolt.new is best for outputting better result in terms of full stack apps ( I am using the free version by the way)
More improvements are comming to open source. I have second most popular fork and we will be joining forces with more people to start iterating on the fork quicker.
@EduardsRuzga I hope cole sees your fork and merges if possible!
@@EduardsRuzga link?
Indeed, I'm looking forward to it!
Do you know of any models or open source apps that focus more on creating better UI's like v0 instead of focusing more on functionality?? That would really help me out.
8:06 this effects are cool😂
What if I decided to put it on a VPS? Do you think that a 16GB, 8 Core VPS will suffice?
Yes
nice work 🎉.
just a question please: how api for chat-gpt4 is free via github? can you explain please? or you mean there is a limit rate?!
thanks once again 🎉
Hello I have tried it on Windows with the Following tasks:
• pip install litellm
• set GITHUB_API_KEY=xxx
• litellm --model github/gpt-4o
but then i get this errors: raise ImportError(f"Missing dependency {e}. Run `pip install 'litellm[proxy]'`")
ImportError: Missing dependency No module named 'backoff'. Run `pip install 'litellm[proxy]'`
Just run.. "pip install litellm[proxy]" and it should probably get fixed.
@@AICodeKing my Link looks like this: Uvicorn running on ←[1m0.0.0.0:4000←[0m (Press CTRL+C to quit)
It's different to yours. When i start and take "OpenAILike" there is no Model...
I hope you know the Answer?
This is the env-File:
# Get your OpenAI Like API Key
OPENAI_LIKE_API_KEY=0.0.0.0:4000
@@AICodeKing Fix this issue by running following commands.
pip install backoff
pip install 'litellm[proxy]'
Thanks @AICodeKing
make this same setup video for windows setup too pleaseee
install wsl on your windows. It works
Which Software have you taken here for Installing? I think it's a software for Linux? But is there any software for Windows too for installing?
Yes, it's the same instructions for windows. Instead of the export command use Setx
Can i build Full Stack Application using this application?
No
my windows not working
You must have install node. Js on windows first
Just Gold!!
By not using Anthropic api key... instead using Mistral etc...isnt that means poor coding output because Claude is the best coding ai assistant superior to Mistral?
Yes,you will get a much better output if you use it with claude 3.5 or gpt-4o,to be honest it is worth the price that you pay for those api’s especially if you need things fast and professionally looking
@@xavierf2229as I watched on the above video, mistral worked better that gpt4o relatively with the free api keys. Do you have any suggestions what would be the better option for a free api key for more professional coding output?
But it doesn't have an Image Upload feature ? 🤔
Good point.
In GitHub repository they mentioned they will update this soon
@@SRan5203yeah I’ve seen it as well! Would be a proper V0 replacement if you feed it some documentation on shadcn…
"Hey, man! Just curious, what’s the Linux distro you’re using here?"
Hope you can have more updates for this
After setting everything up and running it I get "Mistral API Key:Not set (will still work if set in .env file)" does anyone else have this issue, or know how to fix it? Thanks
THE SAME ISSUE HAPPENED HERE
@@yusufisroil Beside where it says "Not set (will still work if set in .env file)" there is a tiny pencil icon, click that and put your API key in there (and make sure your on the Chrome browser as it didn't work for me FireFox) then it should work.
Does this mean its completely free?
Yes
Gemini api is free?? 🤔
Yes. (with rate limits)
Great 👍
Quick question:- can we integrate vortex api(anthropic sonet 3.5 enable api) in bolt.new.if yes then how to do?
Ahora directamente lo puede hacer mediante microsoft copilot, hace 2 días hicieron el anuncio de que ya claude sonnet 3.5 puede correr en VS code
😶🌫bro be dropping out with new bombs every hour nowadays
Hi, First time here! lost me on min 2:22 ..
everything went fine on my OS ,, now ( mv .env.example .env.local ) am stuck lol ,,
i can see the file ( .env.local) any help !! again i have 100% zero skills on coding ... lool ,,
thanks everyone 💚
Just try to rename the env example file to env local and then just proceed to edit it with your api keys.
@@AICodeKingmany thanks,, Still i spend the whole day trying to fiq, out ( .env.local ) on my mac os terminal !!!
@@schoolsat hes using linux so you commands may be different on macos
Any one solved this ?
@@schoolsat ren .env.example .env.local
Just a bit harder to deploy.
thanks bro! That's whats uP
Wow for once i outperformed the master , found this before you
v0 is still better though 100 percent.
i use windows not linux bruh
Love you
Hi AICodeKing, Thanks for another great video. You mentioned a video where you explain how to setup the Github free Models in depth (ua-cam.com/video/p_tyWtQZx48/v-deo.htmlsi=LhgIhTcQUIzFAL3O&t=412) What is the title or link to that video?
This one : ua-cam.com/video/Y4UuCmZFJZo/v-deo.html
Wow
hopefully will add o1 o1-preview soon , tried to add myself not successful
Ollama + HuggingFace