Stage 1 · Lesson 4 of 6
Model families and tiers
Most providers sell a family in three or four sizes. As of October 2026, Anthropic sells Claude Opus, Sonnet and ; OpenAI sells GPT-6 Astra, Sol and Luna; Google sells Gemini Pro, Flash and Flash-Lite. The biggest tier is the most capable and the slowest.
The middle tier is the everyday workhorse. The smallest tier is cheap and fast for simple, high-volume work. Open-weight families such as , Qwen, Gemma, Mistral and DeepSeek come in sizes too, measured in (the number of learned values, such as 27B for 27 billion).
Names change every few months, but the shape stays the same. Learn to ask "which tier does this task need?" instead of memorising names.
A question to check yourself, an AI practice session and 1 cheat sheet entry come with the full roadmap. Unlock
Resources for this stage
Start with· follow one of these
- CoursePaidGrokking Modern AI Fundamentals
Design Gurus · Beginner · $72 one-time
A structured grounding in how models, tokens and embeddings work before you compare them
- GuideFreeApplied LLMs: What We Learned From a Year of Building
Yan, Bartolo, Fu and others · Advanced
Free: practitioners on starting with a strong model, then measuring before you step down