Inference

Tokens per second

The rate at which a model generates output, a common measure of speed.

At 100 tokens per second a 500-word answer takes about seven seconds. Speeds depend on provider load and hardware as well as the model.

All terms

Frequently asked questions

What does Tokens per second mean?

The rate at which a model generates output, a common measure of speed. At 100 tokens per second a 500-word answer takes about seven seconds. Speeds depend on provider load and hardware as well as the model.