onrup

Serving

Temperature

What is temperature?

Temperature controls randomness in token selection. Lower values make output more deterministic and repetitive; higher values make it more varied and more likely to be wrong. For extraction and classification it should be at or near zero.

The most common misconfiguration in production is a default temperature left in place on a task that wanted determinism. Structured extraction at temperature 0.7 produces occasional inventive nonsense for no benefit.

It also makes evaluation harder: a non-zero temperature means the same input can produce different outputs, so results have to be averaged over runs to be comparable.

Related terms

All terms in the glossary →

Start with the free tier

A magic link creates your account, your tenant and your first API key. No card until you ask for compute.