Skip to contents

Start here

The common research tasks, simplest first.

Working with Files and Media

Ask a model about an image, a recording, a video or a PDF page.

Classifying Texts with tidyllm

Label thousands of free-text answers with a codebook and check the labels against a sample coded by people.

Decision Models with tidyllm

Code many texts or pictures with fixed yes/no, choice or score questions and get a probability for each answer, so you read the uncertain ones first.

Embedding Models in tidyllm

Turn texts, images or recordings into numbers, then find similar items, group them or search them.

Structured Question Answering from PDFs

Ask the same questions of a folder of papers and get one table row per paper.

Go deeper

Local models, command line providers, tools and Shiny apps.

Local Models with tidyllm

Run models on your own computer, for confidential data.

Command Line Providers

Chat through a coding agent you already pay for, such as Claude Code, Codex or Vibe, without an API key.

Using Tools with tidyllm

Let a model call your R functions or search the web while it answers.

Using tidyllm in Shiny

Build an app in which many users call models without freezing it.