AI labs are lowering costs by selecting smaller, more efficient models for specific tasks, optimizing inference techniques like caching and batching, and leveraging open-source models as their capabilities improve.
Open the full topic