Phase 3 of 9

LLM Landscape & APIs

Days 21-30 (10 days)

This Is Where It Gets Exciting

Welcome to the heart of the bootcamp. You're about to connect to the most powerful AI systems ever created - GPT-4, Claude, Gemini, and open-source models like Llama.

But first, you need to understand the landscape. Who are the major players? When should you use OpenAI vs Claude vs open source? How much does it cost? This phase answers all of that, then teaches you to build with each provider.

What Problems Will You Solve?

LLM Landscape
Choosing wrong model costs 10x more or gives poor results
OpenAI API
Industry standard - every other API is based on this
Claude API
Often better at coding and writing - need both options
Gemini API
Massive context windows, native video - unique capabilities
Open Source
APIs are expensive, privacy concerns - need alternatives
Function Calling
AI outputs text but apps need structured data to act

Daily Schedule

Day 21

The LLM Landscape Overview

Major players, pricing, open vs closed, leaderboards
Day 22

OpenAI API: Fundamentals

SDK setup, chat completions, roles, temperature
Day 23

OpenAI API: Vision & JSON Mode

Image analysis, structured outputs, reproducibility
Day 24

Anthropic Claude API

Claude SDK, Sonnet vs Opus, XML prompting
Day 25

Google Gemini API

Gemini SDK, 1M context, video understanding
Day 26

Open Source Models via APIs

Llama, Mistral, Groq, Ollama intro
Day 27

Multi-Provider & Router Patterns

Smart routing, fallbacks, cost optimization
Day 28

Embeddings & Semantic Search

Vectors, similarity, foundation for RAG
Day 29

Function Calling & Structured Output

Tool definitions, Pydantic, data extraction
Day 30

Streaming & Token Management

Real-time responses, cost tracking, context limits
← Phase 2 Start Day 21 →