Skip to the content

Guides · 20 pages

Models & cost

Which Claude model to use, what effort levels do, and what it all costs to run.

Models & costThe Claude models, and what each costsFive families, one job each. With the prices, and when the cheap one is the right one.3 min Models & costEffort levels, and what each one costsThe other dial. It changes your bill as much as the model does.3 min Models & costWhich model for which jobA decision guide, not a benchmark table.2 min Models & costWhat a token actually isThe unit you're billed in, and how to estimate it without a calculator.2 min Models & costWhy price comparisons across models misleadNewer models produce about 30% more tokens for the same text.2 min Models & costThe economics of prompt cachingThe biggest free win in model cost, and the arithmetic behind it.2 min Models & costMeasure cost per task, not per tokenThe cheaper model that needs four attempts is the expensive one.2 min Models & costWorking out what a feature will cost to runDo the arithmetic before you build it, not after the first bill.2 min Models & costHalf price for work that can waitThe Batch API is a 50% discount for anything not needed right now.2 min Models & costThe tools that cost extraMost usage is tokens. A few things bill separately.2 min Models & costTool definitions cost tokens on every requestA big tool surface is a standing charge you pay per message.2 min Models & costContext windows, and what fills themA million tokens sounds infinite. It isn't, and it isn't free.2 min Models & costPutting a hard ceiling on your billThe protection that matters most, and the one most often missing.2 min Models & costWorking out where the money wentYour usage dashboard answers questions your intuition won't.2 min Models & costSubscription or pay as you go?Two ways of paying, with different failure modes.2 min Models & costWhen the cheap model is the right modelHaiku and Sonnet do more than people give them credit for.2 min Models & costUsing more than one model in a projectCascades and sub-agents work well. There's one cost you'll miss.2 min Models & costRate limits, and what to do when you hit oneNot a bug, not a bill. A queue.2 min Models & costWhat a really long prompt costsThe million-token window has no surcharge. It still has a price.2 min Models & costWhat a vibe-coded project actually costsWorked examples, from a weekend build to something with users.2 min