Inference is the new cost of goods5 min read · 2026-07-13

When agents loop and sessions grow, the model API bill starts to behave like a cost of goods sold. Here is why the bill has that shape, and what a company can do to hold it down.

One model per prompt no longer fits5 min read · 2026-07-13

Picking a model for a single prompt made sense when work was one message. Agentic work is a session, so you route the whole session and choose where it sits on the cost-quality frontier.

The shape of the AI bill3 min read · 2026-06-18

Agents loop, assistants hold long conversations, and every turn resends the same context. Most of a modern AI bill is the same tokens, paid for again and again.

Where our numbers come from2 min read · 2026-06-12

Every figure we show traces to published provider pricing or to your own traffic, measured.