The Model Landscape for Builders
Frontier, mid-tier, and small models; closed and open-weight; general and specialized — how the market is shaped and how to keep up without chasing every release.
Every week brings a new model announcement, a new leaderboard winner, and a new thread telling you everything you built last month is obsolete. If you follow all of it, you'll spend your time migrating instead of shipping. If you ignore all of it, you'll keep paying frontier prices for work a model twenty times cheaper now does just as well.
This lesson gives you a map that holds still while the names on it change. You'll learn the three axes that shape the model market (capability tier, access, and specialization), what actually differs between models beyond benchmark scores, and a lightweight routine for keeping your product current: a model inventory in code, a script that warns you before a model you depend on is retired, and a rule for when a new release deserves your attention.
Three axes that describe any model
Strip away the marketing and almost every model you might use can be placed on three axes. Placing a model on them tells you more about whether it fits a feature than any single benchmark score.
| Axis | The options | The question it answers |
|---|---|---|
| Capability tier | Frontier, mid-tier, small | How hard a task can it handle, and at what price and speed? |
| Access | Closed (API only) or open-weight (downloadable) | Where can it run, who sees your data, and who controls when it changes? |
| Scope | General-purpose or specialized | Is it a generalist text model, or built for one job like embeddings or speech? |
In Deskmate, drafting a reply to an angry enterprise customer benefits from a strong general model. Tagging 50,000 tickets overnight usually needs only a small one. Turning help articles into search vectors needs an embedding model, a different kind of model entirely. Three features, three spots on the map.
The rest of this module builds on these axes. Choosing a Model per Feature turns them into a selection process, and Model Routing and Cascades shows how to move individual requests between tiers.