Glossary · 2 · The AI landscape: brands and model families
Model selection
Also known as: Choosing an AI model, LLM selection
Model selection is the structured choice of an AI model for a specific use case, weighing task quality, cost per token, speed, context size, openness, hosting and data terms, and regulatory fit — ideally based on tests with your own content rather than on public leaderboards.
- Intermediate
- Expert
- Technical project managers
- Developers
In one sentence
How to choose an AI model: criteria for quality, cost, speed, context, data terms and hosting — and why your own tests beat leaderboards.
Example
A team tests three models on 50 real customer questions against its documentation, scores the answers and picks the second-best model because it is five times cheaper and nearly as accurate.
Why it matters on your learning path
- Technical project managers: Document the criteria and results; the choice will be revisited when new models appear, often within months.
- Developers: Build an evaluation set once and rerun it for each candidate model; abstract the provider behind an interface.
Criteria checklist
- Quality on your tasks and languages
- Cost per request at expected volume
- Latency for interactive use
- Context window and multimodal needs
- Data terms: training on your data, retention, region — see data privacy in AI tools
- Openness: API only or open weights
- Governance: documentation such as model cards and provider obligations under the EU AI Act