Story

In brief

How fast is 10 tokens per second really? Neat little HTML app by Mike Veerman ( source code here ) which simulates LLM token output speeds from 5/second to 800/second. Useful if you see a model advertised as "30 token...