Guides

Turn on prompt caching by adding two lines to a request, and prove from the response's own numbers that it worked.

Define one tool, send it with a question, and see the model ask for it instead of guessing — in a file you wrote and ran yourself.

Write an MCP server that offers one tool, and send it a message by hand to see exactly what it answers.

Write a JSON Schema for the data you need, and send it with a request so the reply follows it.

Read how much memory your machine has, and how much of it belongs to the graphics card, with one file you run yourself.

List the models you want to run and say how much memory each one needs, with one file you run yourself.