AI is TOO EXPENSIVE if You Don't Do This

Artificial Intelligence is no doubt the future of not just software development but the whole world. And I'm on a mission to master it - focusing first on mastering AI Agents.
In this video I show you how to use AI Agents to save a TON of money with LLMs. It's all about using an AI Router (a type of agent) to route simple requests to a cheaper model and more complex requests to a more powerful and expensive model. I even show exactly how to implement this using Llama 3 70-B (powered by Groq) and GPT-4o!
All the code created in this video can be found here:
github.com/col...
Join me as I push the limits of what is possible with AI. I'll be uploading videos twice a week - Sundays and Wednesdays at 7:00 PM CDT!

[@chadpogs7973]

Wow!! I found a gem!! Great content! Keep it coming!!

[@ColeMedin]

@@chadpogs7973 Thank you very much - I appreciate it a ton! I'll definitely keep the content coming 👍

[@stonedizzleful]

Great video!

[@yazanrisheh5127]

I'm a bit confused. Do you first use an LLM to determine if its a complex or simple question and then you use 2 different LLMs depending on their complexity?

[@ColeMedin]

Yes that is exactly right! You use a cheaper model (like Llama 3.1) to determine the complexity of the prompt. Then, if the prompt is deemed more complex, you send it to a more powerful model (like GPT-4o). And if it's a simpler prompt you will send it to a cheaper model (like Llama 3.1, it can be the same as the first model!).

[@lsd7777777]

Hi. It would be great to see this implementation in n8n

[@ColeMedin]

@@lsd7777777 I appreciate the suggestion! I'll consider doing that for sure!

[@424kchun424]

@@ColeMedin also for both cheap and expensive AI to share same memory context consistently

[@ColeMedin]

@@424kchun424 Great point! Yes, that will be super important for sure and I will keep that in mind!