Which AI Model Do I Choose?

In this practical session, Microsoft MVP Tine Staric explains what separates small, large, and reasoning models, and how to choose the right AI model based on capability, cost, response time, and data privacy.

Choosing the right AI model can make a significant difference to your GitHub Copilot costs. As more colleagues adopt AI coding assistants and agents, model choices that once went unnoticed can become expensive habits.

In this practical session, Microsoft MVP Tine Staric follows a prompt from submission to response, explaining what actually gets sent to the model and how caching influences what you pay. Discover what separates small, large, and reasoning models, when additional thinking is worthwhile, and why price per token tells only part of the story.

Learn to evaluate AI models by the cost of completing a task and identify when a premium model delivers meaningful value. Tine also explores whether open-weight models offer a practical alternative, where your data goes when another provider hosts them, and what running a model on your own hardware involves.

Leave with a clearer understanding of which tasks suit which models and how to make informed choices as AI usage grows across your team.