Napkins + Llama-3.2 : Generate Apps from Screenshots for FREE with Llama-3.2 90B!

Join this channel to get access to perks:

---
In this video, I'll be telling you that how you can generate Full-Stack Production-Ready Web Applications using AI with Napkins and Llama-3.2 (90B). I'll be creating a full-stack App using AI / Llama-3.2 90B without writing any code and for 100% FREE. This going to be a full-hands on about how you can create full-stack applications with AI. I have chosen NextJS as my frontend and Supabase as my backend and I'll be creating a full-stack app without writing any code. You can also use this with other Opensource LLMs like Gemma-2. DeepSeek-Coder-V2, Qwen2, Llama-3 and other such LLMs.

-----
Key Takeaways:

🔥 Llama-3.2 launched, but is it good enough? Check out my testing results and how it compares to expectations in real-world AI use cases.

🚀 Meta & Together AI’s "Napkins" app transforms wireframes or pictures into React and Tailwind apps for free—get started with powerful AI-generated code now!

💡 Llama-3.2 Vision models power Napkins, but can they accurately convert designs to apps? Watch to see if it’s the coding solution you've been looking for.

💻 Tested both 90b and 11b models—find out which model performs better in converting wireframes into usable code with minimal effort!

📲 How to host Napkins locally for ultimate control! Follow my step-by-step guide to run this free AI tool on your own machine with the Together API.

🛠️ Is Napkins the best free tool? Explore its strengths and limitations, and learn whether it’s the right choice for building your next app prototype.

🎨 Simple landing page generation made easy! See how Napkins can turn text prompts into stunning app designs with just a few clicks—perfect for quick mockups!

----
Timestamps:

00:00 - Introduction
00:07 - About Llama-3.2
00:23 - About Napkins based on Llama-3.2 (Screenshot to App)
01:08 - Usage & Testing
05:26 - Local Installation
07:38 - Ending

[dejohnny]

Fail, fail, fail. I won't be using Napkins.

[jmarkinman]

Maybe have it pop out a text description instead of code and send the text description to 3.1 70B or 405B to print the code?

[doctor]

Definitely need an update when new models are added.

[josemarques]

Is it possible to change the model used by Napkins? I thought about changing the Napkins code to use Qwen 2 instead of Llama 3.2.

[DouhaveaBugatti]

I think i should play with it a little bit 😅

[springou]

Hello~ May I invite you to review my AI video/image enhancer software?

[lokeshart]

Btw which is the best ai for responsiveness in html css and js coding 😮

[voiceall]

What are the advantages of using it locally if it is free on the cloud already? Are you suggesting to add fastapi and make api call?