Stage 2 · Lesson 8 of 9
Hosted vs local models
APIs give you the strongest models with no hardware to manage, but you pay per token and send your data to a provider. are ones whose trained weights (the model files) anyone can download, such as the , Qwen, Gemma or families.
They can run on your laptop through Ollama, which serves an at localhost:11434. Local models are ideal for free experiments, offline work, tests and data that must not leave the machine. But a small model on a laptop is slower and noticeably weaker than a frontier hosted model (one of the strongest current models).
Put the model behind one small function, or behind a provider-agnostic SDK that works with many providers. Then you can switch with a config change and compare models on the same inputs.
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 Prompt Engineering for Software Engineers
Design Gurus · Beginner · Subscription
Prompting patterns for engineers building features, not for chatting
- ToolFreeVercel AI SDK Documentation
Vercel · Intermediate
Free: the Foundations and Getting Started guides for prompts, generateText and streamText
Practise
Keep handy
Also worth your time
Few-shot, chain of thought, JSON enforcement and parsing failures in one place