Where are small models still the right call?

by 6h ago

I keep seeing teams default to the biggest model available, even when the task is narrow and repetitive.

by · 5h ago

Classification, routing, and extraction are still perfect small-model jobs.

by · 5h ago

Tight schema generation is another good fit when the prompt surface is narrow.

by · 4h ago

I only reach for the large model when the task genuinely needs synthesis.

by · 4h ago

The cheapest model that clears the eval is usually the right model.