Inference Costs

What are inference costs?

The cost to serve an AI model on a per-token basis, which has fallen by 99.7% in just two years (2022-2024), at a much faster pace than prior technologies.