๐Ÿค–

Intro to LLMs

๐Ÿง‘โ€๐ŸŽ“ Apprentice

What is an LLM?

LLM = Large Language Model

It's a program trained on millions of texts that can understand and generate human language. Including code.

The key insight: An LLM can write code for you. You just have to ask properly.


The "Prompt-First" concept

Instead of memorizing syntax and copying code from Stack Overflow, you can:

  1. Describe what you want in natural language
  2. Ask the LLM to write the code
  3. Run the code it gives you
  4. Iterate if something doesn't work

This works with any LLM: Gemini, ChatGPT, Claude, etc.


Where can I use an LLM?

OptionAccessCostIdeal for
Google AI StudioGoogle accountFreeGetting started
ChatGPTOpenAI accountFree/PaidGeneral use
Claude.aiAnthropic accountFree/PaidCode and reasoning
OllamaYour computerFreeTotal privacy

๐Ÿ’ก Recommendation: Start with Google AI Studio. It's free, no credit card, and you already have an account.


Main LLMs (2026)

ModelCompanyStrength
GeminiGoogleFree, multimodal, easy API
ClaudeAnthropicExcellent for code, reasoning
GPT-4OpenAIVersatile, broad knowledge
LlamaMetaOpen source, runs locally

What is an API Key?

When you use an LLM from your code (not from the web chat), you need an API Key.

It's like a password that identifies who is using the service.

โš ๏ธ Never share your API Key or upload it to GitHub.


Key concepts

ConceptWhat it isExample
PromptWhat you ask the model"Write a chatbot in Node.js"
ResponseWhat the model answersThe code + explanation
TokenUnit of text (~4 characters)"Hello" = 1 token
ContextWhat the model "remembers"Previous conversation

Practice

โ†’ Chatbot with Gemini โ€” Your first AI project


Useful links