Guide · updated October 13, 2026
Connect the pieces
The standing explanation. Today’s briefing is what changed. This is how the pieces fit, what they cost, and how a team gets better at one job.
Agents, loops, and tools
A model writes the next tokens. A tool does something in the world. A loop decides whether to go again. An agent is those three, aimed at one goal.
How models are ranked, and what they cost
A rank is a test with a name. A price is a list rate per million tokens. Neither one is a verdict on your job.
Fine-tuning for one job
Fine-tuning changes a model you already have, using examples of a decision your team can score. It is how a company owns a narrow task without training a foundation model.