Our LLM cost model was double-counting batch and cache discounts

 (beforeyouship.dev)

1 points | by indiegoing 4 hours ago ago

1 comments